[go: up one dir, main page]

WO2018006065A1 - Procédés et systèmes pour modifier l'influence de l'utilisateur pendant une session collaborative d'un système d'intelligence collective en temps réel - Google Patents

Procédés et systèmes pour modifier l'influence de l'utilisateur pendant une session collaborative d'un système d'intelligence collective en temps réel Download PDF

Info

Publication number
WO2018006065A1
WO2018006065A1 PCT/US2017/040480 US2017040480W WO2018006065A1 WO 2018006065 A1 WO2018006065 A1 WO 2018006065A1 US 2017040480 W US2017040480 W US 2017040480W WO 2018006065 A1 WO2018006065 A1 WO 2018006065A1
Authority
WO
WIPO (PCT)
Prior art keywords
user
users
pointer
time
group
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/US2017/040480
Other languages
English (en)
Inventor
Louis B. Rosenberg
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Unanimous AI Inc
Original Assignee
Unanimous AI Inc
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from PCT/US2016/040600 external-priority patent/WO2017004475A1/fr
Priority claimed from US15/199,990 external-priority patent/US20160314527A1/en
Priority claimed from US15/241,340 external-priority patent/US10222961B2/en
Application filed by Unanimous AI Inc filed Critical Unanimous AI Inc
Priority to US15/815,579 priority Critical patent/US10439836B2/en
Priority to PCT/US2017/062095 priority patent/WO2018094105A2/fr
Publication of WO2018006065A1 publication Critical patent/WO2018006065A1/fr
Priority to US15/898,468 priority patent/US10712929B2/en
Priority to US15/904,239 priority patent/US10416666B2/en
Priority to US15/922,453 priority patent/US20180204184A1/en
Priority to US16/059,698 priority patent/US11151460B2/en
Priority to US16/154,613 priority patent/US11269502B2/en
Priority to US16/230,759 priority patent/US10817158B2/en
Anticipated expiration legal-status Critical
Priority to US17/237,972 priority patent/US11636351B2/en
Priority to US18/114,954 priority patent/US12099936B2/en
Ceased legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G09—EDUCATION; CRYPTOGRAPHY; DISPLAY; ADVERTISING; SEALS
    • G09B—EDUCATIONAL OR DEMONSTRATION APPLIANCES; APPLIANCES FOR TEACHING, OR COMMUNICATING WITH, THE BLIND, DEAF OR MUTE; MODELS; PLANETARIA; GLOBES; MAPS; DIAGRAMS
    • G09B7/00—Electrically-operated teaching apparatus or devices working with questions and answers
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/004—Artificial life, i.e. computing arrangements simulating life
    • G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/10—Office automation; Time management
    • G06Q10/101—Collaborative creation, e.g. joint development of products or services

Definitions

  • the present invention relates generally to systems and methods for collaborative intelligence, and more specifically to systems and methods for closed-loop, dynamic collaborative intelligence. Even more
  • the invention related to analysis methods for decisions made by a collaborative intelligence comprised of networked user working as a distributed real-time closed-loop dynamics system.
  • Portable computing devices such as cell phones, personal digital assistants, and portable media players have become popular personal, devices due to their highly portable nature, their ability to provide accessibility to a large library of stored media files, their
  • the current art has some deficiencies.
  • One deficiency is the fact that participating users in a real-time closed-loop collective intelligence systems may sometimes be excessively influenced by other users i the system, rather than expressing their own individual will, wisdom, or intent. This is known as “social influence bias” and has been shown to have negative effects on the collective intelligence of human groups. Human swarms are more resistant to "social influence bias” than traditional polls, surveys, and markets, as a result of their highly parallel nature, but there is still some social biasing effects to be addressed.
  • the present invention aims to reduce social biasing effects in real ⁇ time closed-loop collective intelligence systems.
  • a related deficiency of the present art is that participants within human swarms sometimes do not convey their individual conviction level with respect to a set of options currently under group deliberation with enough expressiveness to clearly differentiate the varying levels of conviction among users. What is therefore needed are predictive methods that more expressively reflect the conviction levels of use s as they vie for particular options within a real-time closed-loop
  • a collaboration system including a plurality of individual computing devices and a collaboration server exchanging data with each of the plurality of computing devices, wherein the user is included in a group of individual users
  • each user of the group of individual users associated with and interacting with one of the individual computing devices, wherein each user is associated with one user intent vector representing a magnitude and direction of input of that user in real- time, 'wherein past magnitudes during the session comprise a magnitude history and past directions during the session comprise a direction history, whereby the combined user intent vectors are used to determine movement of a displayed graphical pointer used to select one of the input choices, comprising the steps of:
  • the user input vector determining, based on the direction history, how many times during the session the user input vector changed from a direction towards one of the input choices to a direction towards a different input choice; and changing, upon determining in real-time that the user input vector has changed from a direction towards a first input choice towards a second input choice, the magnitude of the user intent vector .
  • the invention can be any organic compound.
  • the invention can be any organic compound.
  • a method for adjusting a user intent vector of a user during a collaboration session for a collaboration system including a plurality of individual computing devices and a collaboration server exchanging data with each of the plurality of computing devices, wherein the user is included in a group of individual users participating in selection of one of a plurality of displayed input choices, each user of the group of individual users associated with and interacting with one of the individual computing devices, 'wherein each user is associated with one user intent vector representing a magnitude and direction of input of that user in real- time, wherein the plurality of user intent vectors is used to determine a group intent vector representing a direction and magnitude of input of the group of user in real-time, comprising the steps of: determining, based on the user intent, vector, a current user-selected input choice of the plurality of input choices; determining if the current user-selected input choice is different from a previous user-selected input choice; comparing, upon determining that the current user-selected input choice is different from the previous user-selected
  • FIG. 1 is a schematic diagram of an exemplary real ⁇ time collaborative system..
  • FIG. 2 is an exemplary display interface of a computing device of the collaborative system in
  • FIG. 3 is an exemplary group display interface of the computing device of the collaborative system at a point in time during a collaboration session.
  • FIG. 4 is an exemplary group display interface of the computing device of the collaborative system after the collaboration, session has been successfully
  • FIG. 5 is a frame of an exemplary collaboration session replay video
  • FIG. 6 is an exemplary display interface during a collaboration session determining whether to e ect a specific member from the group.
  • FIG. 7 is an exemplary display interface during a session determining whether to allow a specific member to j o in the group .
  • FIG. 8 is an example display interface of the virtual lobby interface.
  • FIG. 9 is an exemplary first time step target area of an exemplary display interface shown at a first time step of an exemplary decision period.
  • FIG. 10 is an exemplary second time step target area of the exemplary display interface shown at a second time step of the exemplary decision period.
  • FIG. 11 is an exemplary third time step target area of the exemplary display interface shown at a third time step of the exemplary decision period.
  • FIG. 12 is a flowchart diagram of a method of faction determination in accordance with another
  • FIG. 13 is a plurality of time step target areas showing faction-associated spatial regions for three different time steps of the exemplary decision period, in accordance with a first embodiment of the method of faction determination of FIG. 12.
  • FIG. 14 is a plurality of time step target areas showing faction-associated spatial regions for three different time steps of the exemplary decision period, in accordance with a second embodiment of the method of faction determination of FIG. 12.
  • FIG. 15 is a spatial region diagram of a portion of the first time step target area as shown in accordance with the second embodiment of the method of FIG. 12.
  • FIG. 16 is a spatial region diagram of a portion of the second time step target area as shown in accordance with the second embodiment of the method of FIG. 12.
  • FIG. 17 is a snapshot of an exemplary target area during an exemplary collective intelligence decision process .
  • FIG. 18 is an exemplary time-history plot depicting a time-history of faction data across the decision period of the collective intelligence decision of FIG. 17.
  • FIG. 19 is a flowchart diagram, of a first method for adjusting the pull strength of a user.
  • FIG. 20 is a flowchart diagram of a second method for adjusting the pull strength of a user.
  • Real-time occurrences as referenced herein are those that are substantially current within the context of human perception and reaction.
  • the massive connectivity provided by the Internet is used to create a real-time closed-loop collaborative consciousness (or emergent group-wise intelligence) by collecting real-time input from large numbers of people through a novel user interface and processing the collected input from that large number of users into a singular group intent that can collectively answer questions, make statements, take actions, select functions, or otherwise respond to prompts in real time.
  • the methods use intervening software and hardware to moderate the process, closing the loop around the disparate input from each of the many individual participants and the singular output of the group .
  • a collaboration system has been developed that allows the group of users to collaboratively control a graphical pointer 210 in order to collaboratively answer questions or otherwise respond to prompts.
  • each individual user (“participant”) engages the user interface on a computing device 104, conveying his or her individual real-time will in response to a prompt such as a textually displayed (or audibly
  • a "social swarming" platform is enabled that allows users to join one of a plurality of hosted groups (also referred to as swarms), each group comprising a plurality of users.
  • the user may collaborate with that group, earn scores and/or credits and/or rankings based on his performance with respect to others in the group, and browse the stored output from other groups .
  • groups can compete with other groups, each of said groups also earning group scores, credits, and/or rankings with respect to other groups.
  • FIG. 1 a schematic diagram of an exemplary collaboration system 100 is shown. Shown are a Central Collaboration Server (CCS) 102, the plurality of portable computing devices 104, and a plurality of exchanges of data 106 with the Central Collaboration Server 102.
  • CCS Central Collaboration Server
  • Embodiments of the plurality of portable computing devices 104 and the interaction of the computing devices 104 with the system 100 are previously disclosed in the related patent applications.
  • the system. 100 comprises the Central Collaboration Server (CCS) 102 in communication with the plurality of computing devices 104, each of said computing devices 104 running a Collaborative Intent Application ("CIA") .
  • the system 100 is designed to enable the plurality of users, each engaging an interface of one of said computing devices 104, to jointly control a single graphical element, for example the movable pointer 210, through real-time group-wise collaboration.
  • the portable computing devices 104 may communicate 'with each other.
  • the CCS 102 includes software and additional elements as necessary to perform the required functions. In this application, it will be understood that the term.
  • CCS may be used to refer to the software of the CCS 102 or other elements of the CCS 102 that are performing the given function.
  • multiple pointers controlled by multiple swarms also referred to as groups
  • cur ent discussion we will give examples that are confined to a single swarm. This is for simplicity of description and is not intended to limit the scope of the innovations.
  • each of the computing devices 104 comprises one or more processors capable of running the CIA routines and displaying a representation of the pointer 210 along with a plurality of graphical input choices 208.
  • the computing device 104 could be, for example, a personal computer running the CIA
  • the CIA software code can be configured as a stand-alone
  • executable or be code that executes inside a web-browser or other shell.
  • FIG. 1 shows only six computing devices 104 in communication 'with the CCS 102
  • the system 100 is highly scalable, enabling hundreds, thousands, or even millions of users to connect simultaneously to the CCS 102, each using their own computing device 104, thereby sharing a real-time collaborative experience with the other users. In this way, large numbers of users can collaboratively control the pointer 210 to generate a response as a group inte11 igence .
  • FIG. 1 shows simple top-down architecture for direct communication between the CCS 102 and each of the computing devices 104
  • related application 14/708,038 entitled MULTI-GROUP METHODS AND SYSTEMS FOR REAL-TIME
  • MULTI-TIER COLLABORATIVE INTELLIGENCE discloses multi- group and tiered architectures that enable shared
  • each of said computing devices 104 that is engaged by a participating user includes one or more display devices for presenting a graphical user interface to the user.
  • an exemplary display interface 200 is shown in accordance with one embodiment of the present invention. Shown are a prompt bar 202, a group name 204, a target area 206, a plurality of input choices 208, the pointer 210, a communication menu 212, a board selection drop-down menu 214, a physics selection drop-down menu 216, a chat window 218, a chat input box 220, a current member list 222, a statistics display 224, an invite button 226, and an ask button 228.
  • the collectively controlled graphical pointer 210 is simultaneously displayed to each user by the CIA running on his computing device 104.
  • the pointer 210 displayed to each user appears in a substantially similar position with respect to a set of input choices 208 (as compared to the relative position of the pointer 210 and input choices 208 on other user's screens) .
  • the synchrony of the interfaces is coordinated by the data 106 received by each computing device 104 sent from the CCS 102 over the communications link.
  • data 106 is sent from the CCS 102 to each of the plurality of computing devices 104 at a rate of 60 updates per second, the data 106 including the current position of the graphical pointer 210 (also referred to as a puck) with respect to the set of graphical input choices 208, as further shown below.
  • Coordination data may also include orientation information .
  • the input choices 208 are identically displayed upon all the computing devices 104, although some unique embodiments allow for divergent input choices 208.
  • the input choices 208 are displayed in the native language of each user, each input choice 208 conveying a substantially similar verbal message, but translated based on a language setting of the user. This feature enables swarms of individuals who may speak different languages and are unable to communicate directly, to still form a
  • the displayed questions are also automatically translated into the chosen native language of the user. This is also true of a displayed answer, and optionally the chat window 218 output.
  • multiple graphical pointers 210 are displayed by the computing devices 104, each of said graphical pointers 210 being collaboratively controlled by a different group of users.
  • 500 users may be collaboratively controlling Graphical Pointer #1, while a different group of 500 users are collaboratively controlling Graphical Pointer #2.
  • the first group of 500 users comprises a first collective intelligence.
  • the second group of 500 users comprises a second collective intelligence. This unique system and methods allow for the first collective intelligence to compete with the second collective intelligence in a task that is
  • one collective intelligence can be enabled to complete with another collective intelligence in a real-time trivial
  • the CIA software running on each computing device 104 is configured to display a graphical display interface 200 that includes at least one
  • the graphical pointer 210 is configured to look like a "glass puck” with a central viewing area that is transpare t.
  • the input choices 208 are configured as a hexagon of six input choices 208, each input choice 208 including a graphical icon (in the embodiment shown, a dot inside a circle) and an associated word.
  • the six input choices 208 correspond with possible answers to questions: "Yes”, “Maybe”, “No”, “Yes", “Bad Question", and "No".
  • the pointer 210 When the pointer 210 is positioned over one of the input choices 208 such that the input choice 208 is substantially within a centralized viewing area of the pointer 210 for more than a threshold amount of time, that input choice 208 is selected as a target.
  • the threshold amount of time is 3 to 5 seconds.
  • the centralized viewing area appears as a graphical etching on the glass pointer 210, the etching remaining invisible until the pointer 210 approaches a target.
  • the spatially arranged graphical input choices 208 can.
  • the input choices 208 could also comprise
  • the user enters the question into the prompt bar 202, Once entered, the user clicks the ask button 228, 'which sends the question from, the CIA software of that particular user (running on his local computing device 104) to the CCS 102. Because many users could ask questions, the CCS 102 acts as the gate keeper, deeming the first question received, (when no question is currently in process) as the one that will be asked to the group. In the current embodiment, not all users are enabled to ask questions at any given time to avoid too much competition for asking. In some embodiments, credits are redeemable by the user for the right to ask the question. In some embodiments, the user must spend credits to ask the question, and can only ask if he has enough credits . In some embodiments, users earn credits based on points awarded for participation in a session. More credits are awarded to users who have high
  • the methods for computing sync scores will be described in more detail further below .
  • users can select from a plurality of possible target boards by using the board selection drop-down menu 214.
  • the currently selected target board is for yes/no questions.
  • Other target boards may include true/false questions, good/bad ques ions, and o he sets of standardized answers.
  • a spelling board may be included where a full alphabet of input choices 208 are displayed, allowing users to spell out answers (as shown in co-pending applications) .
  • the spelling board may also include numbers, punctuation, backspace, blank space, and other alphanumeric
  • boards can also be entered by selecting "custom” from, the board selection drop-down menu 214.
  • “suggestion mode” can also be selected for a given question through the board selection drop-down menu 214, which, asks other users in the plurality of users to give s gges ions that populate the board in real-time .
  • users can selectively use a physics mode from, the physics selection drop-down menu 216.
  • a physics mode has been selected, but users can choose ice mode where the pointer 210 slides around, on the target board, as if it were
  • A. gravity mode is configured, to pull the pointer 210 back, to a center of the input choice set (i.e. center screen) as if by simulated gravity.
  • a heavy mode the pointer 210 has substantially higher mass than in standard mode and. thus is harder for users to move.
  • a barrier mode a set of physical barriers block a direct path to the input choices 208, forcing users to collaboratively guide the pointer 210 around barriers to reach the input choices 208.
  • the display interface 200 includes the chat window 218 that allows users to
  • chat input box 220 Also included is the list of current members who are part of the group and thus enabled to ask questions and collaboratively provide cont ol over the poi ter 210.
  • this group display interface 200 Because users enter this group display interface 200 from a lobby display interface where the user can choose from among a plurality of available collective
  • the name of the current collecti e intelligence group (swarm) is also displayed.
  • users can invite their friends to this group by clicking on the invite button 226 includes in the communication menu 212.
  • these invites can leverage existing social networks such as
  • statistics display 224 that gives the user of this instance of the software (on this computing device 104) a listing of his personal statistics including his score, credits, performance value, the number of rounds he has participated in, and the number of questions he has asked the collective intelligence group.
  • the question is sent by the CIA on that user's computing device 104 to the CCS 102. If the CCS 102 software determines that the question is valid, the question is then sent to all the users in the group so that it appears substantially simultaneously on the display interface of each of the computing devices 104.
  • the question appears in a large box at the top of the target board. Then a "3" - “2" - “1" countdown timer appears at the center of the target board, notifying users get ready for the collaborative answer process, or session, to begin.
  • the display interface (having received instructions from the CCS 102) then displays a graphical "GO" and the users will then collaboratively control the motion of the pointer 210, guiding it towards whichever input choice 208 best satisfies the collaborative will of the group as emergent from the real-time collective intelligence.
  • the collaborative control may be implemented by each user imparting a real-time intent regarding a desired motion of the puck by manipulating a graphical magnet icon 306 on his or her local computing device.
  • the graphical magnet icon 306 defines a magnitude and direction, of the user's pe sonal intent, referred to herein as a user inte t vector.
  • Each answer session is generally limited in total time by the underlying software of the present system. 100, for example giving the swarm 60 seconds to converge upon an answer through the collaborative motion of the pointer 210. This time pressure is deliberate, for it inspires users to employ their gut instincts and
  • the countdown clock 304 is displayed on a group display interface 300 of each user (as shown below in FIG, 3 ⁇ , the timing of the plurality of countdown clocks 304 coordinated by handshaking signals from the CCS 102. If the pointer 210 does not reach the target within the allotted 60 seconds, the system 100 determines that the collaboration is a failure, and sends a failure indication to the CIA of each computing device 104. In some embodiments, in response to receiving the failure indication the CIA terminating user input and displaying the words "brain freeze! on the group interface. In addition, in response to receiving the failure indication all users could lose a number of points and/or credits for the collective failure of the group to guide the pointer 210 to a target .
  • the system 100 is configured to determine that a target is achieved when the group successfully positions the pointer 210 over one input choice 208 for more than the threshold period of time.
  • the target is displayed on the CIA screens of all the users as the answer to the question.
  • statistics for that answer as shown below in FIG. 4, such as the group cohesiveness score and the user synchronicity value, as previously described in related application 14/708,038.
  • Also displayed may be points and/or credits awarded for the current user' s participation in the emergent answer, as shown in FIG. 4.
  • FIG. 3 shown is the exemplary group display interface 300 of one user at a point in time during a collaboration session, i.e. after the questio has been received by the computing devices 104 but before the collaboration session has ended. Shown are the group name 204, the target area 206, the plurality of input choices 208, the pointer 210, the communication menu 212, the chat window 218, the chat input box 220, the current member list 222, the statistics display 224, the invite button 226, a question display 302, a
  • the basic layout of the display interface 300 is similar to FIG. 2.
  • the prompt bar 202, the ask button 228, the board selection drop-down menu 214, and the physics selection drop-down menu 216 have been replaced by the question display 302,
  • the question display 302 appears substantially simultaneously upon the screens of the computers of all users in the swarm. Also displayed on the target area 206 are the set of input choices 208 from which the users are being asked to collaboratively select from.
  • the question is - "What movie should we see tonight?” and the input choices 208 include five movie names: “Jaws”, “Gremlins”, “Stand By Me”, “Indiana Jones”, and “Twister” along with a sixth input choice 208, "Bad Question”.
  • the Bad Question the Bad Question
  • Question choice is automatically included in the input choices 208 by the CCS 102, allowing the swarm to
  • the "3" ⁇ "2"-- “1" countdown timer appears (not shown) to signal the start of the current session.
  • the users are now enabled to provide user input to the pointer 210, guiding it towards one of the input choices 208.
  • the 60 second countdown clock 304 counts down, applying time pressure to the group.
  • the countdown clock 304 is shown at 0:51, indicating that 51 seconds remain in the current session.
  • group members may also be inputting messages via text using the chat window 218, and/or may be chatting with a simultaneously enabled group voice chat. This allows i terpersona1 comniumicat ion during the ses s ion .
  • each user is enabled to apply forces upon the pointer 210 to convey his individual intent as to how the pointer 210 should move at any moment in time.
  • the displayed motion of the pointer 210 is not a reflection of that user's individual input but a reflection of the collectively combined group input from the entire swarm of users.
  • the collective input from the plurality of users can be such that each user' s input imparts an equally weighted contribution to the total force applied to the pointer 210.
  • weighting factors are used to give the input force from, some users a higher contribution as compared to other users.
  • each user is enabled to apply forces upon the pointer 210 using one of a variety of innovative methods.
  • each user is enabled to apply forces upon the pointer 210 using one of a variety of innovative methods.
  • each user controls the graphical magnet icon 306 by
  • a mouse, touchpad, touchscreen, tilt interface, or other provided user-interface method manipulating a mouse, touchpad, touchscreen, tilt interface, or other provided user-interface method.
  • the user moves his mouse cursor within a threshold distance of the pointer 210, it turns into the magnet icon 306 that grows larger in size, the closer to the pointer 210 the mouse is positioned.
  • the larger size indicates a larger force.
  • the relative position of the magnet icon 306, which always orients itself towards a center of the pointer 210 under software control, indicates the direction of pull that user wants to impart on the pointer 210. In this way, a user can intuitively impart of force of a selectable magnitude and direction upon the pointer 210.
  • the user can tilt the portable computing device 104 to convey a desired
  • the magnet icon 306 or other graphical indicator is displayed to indicate the imparted force.
  • the user must also tap the screen while tilting the computing device 104, the frequency of the taps causing a higher force to be applied. This unique use of a
  • the combined tilt and tap methodology is highly effective, for it enables one handed input from users o small mobile devices. It also enables the ease of tilting, but avoids it feeling too passive by also requiring frequent tapping. In many such embodiments, the maximum force is applied for only a short time following each tap (for example 0.5 seconds) and then fades away over a
  • the displayed magnet icon 306 shrinks and fades away along with the force magnitude. This is a highly
  • the user is enabled to swipe across a touchscreen display to indicate the magnitude and direction of the force the user desires to apply to the pointer 210.
  • the magnet icon 306 is displayed, indicative of the magnitude and direction conveyed by the swipe.
  • the swipe force is applied for only a short time (for example 0.5 seconds) and then fades away over a period of time (for example 3 to 5 seconds) .
  • the magnet shrinks and fades away along with the force magnitude. This is a highly intuitive interface and requires that the user repeatedly swipe the screen to maintain a maximally applied force upon the pointer 210. This is an
  • the CCS 102 software collects input from the plurality of users, computes a resultant motion of the pointer 210, and communicates the resultant motion of the pointer 210 to each CIA of the plurality of computing devices 104.
  • the CCS 102 software also determines if the pointer 210 location is successfully targeting one input choice 208 for more than the threshold amount of time. If so, the CCS 102 software determines that the question is answered and communicates the targeted input choice 208 to all members of the group such that it is substantially simultaneously displayed upon the display interface of each computing device 104 included in the group. In this way, the system 100 of the present invention enables groups of networked users to collaboratively control the graphical pointer 210 in response to one or more questions posed by members of group. More
  • embodiments of the current system 100 enable each of the plurality of users to view on a screen of their own individual computing devices 104, a
  • the user intent is represented as a user intent vector.
  • the user intent vector can be conveyed by the user, for example, by tilting his computing device 104 in the desired direction, swiping the screen in a desired direction, or positioning rhe mouse such that the
  • graphical magnet icon 306 pulls on the pointer 210 with a desired direction.
  • eye tracking hardware and software are included in the computing device 104, for example the eye tracking hardware and software disclosed in U.S. Patent No. 7,429,108 to the present inventor.
  • the CIA is configured to operate the eye tracking hardware and software and receive input from the eye tracking hardware are software.
  • a user' s gaze is tracked by the CIA and used to compute the user intent, vector that represents the user's desired motion of the pointer 210, which is communicated to the CCS 102 software. More specifically, the user's gaze defines a location with respect to the pointer 210. The vector between the location and the center of the pointer 210 is then, used by the CIA to compute the magnitude and direction of the user intent vector.
  • the user can simply look towards a direction that he desires the pointer 210 to move, and the user intent vector is computed by the CIA and sent to the CCS 102 software by the CIA.
  • the magnet icon 306 or other graphical element is displayed to represent the user- intent vector on the display. In this way, the user can participate in the collaborative swarm intelligence experience using a hands-free method.
  • a brain-computer interface (sometimes called a mind-machine interface, direct neural interface, synthetic telepathy interface, or a brain- machine interface) , is employed to collect the user input of one or more users in the swarm.
  • a mind-machine interface sometimes called a mind-machine interface, direct neural interface, synthetic telepathy interface, or a brain- machine interface
  • a brain-computer interface is employed to collect the user input of one or more users in the swarm.
  • the user's brain-waves are detected by the brain-computer interface as he or she watches the pointer 210 move upon, his screen.
  • a calibration session is often required to correlate detected brain activity with a desired direction of motion of the pointer 210, but once that calibration is complete, the brain-computer
  • the interface system can be used by the CIA to compute the user intent vector that represents that user's desired motio of the pointer 210 at each time-step during the session, this user intent vector being communicated to the CCS 102 software. In this way, the user can simply think about a direction that he desires the pointer 210 to move, and the user intent vector is computed and sent to the CCS 102 software by the CIA.
  • the magnet icon 306 or other graphical element is
  • the user can participate in the collaborative swarm, intelligence using a hands-free method .
  • the system is configured such that the user intent vector is
  • the CCS 102 collects the user intent vectors from the
  • the group intent vector is then used to compute an
  • a physical model is employed in which the pointer 210 is assigned a simulated mass and damping, each user input represented as a simulated force vector. In some such embodiments, the mass and damping of the pointer 210 is adjusted
  • the ice mode can be selected by the user in which the pointer 210 glides very freely as if on ice.
  • the heavy mode can be selected by the user in which the pointer 210 requires the collaborative pull of a large majority of members of the swarm to achieve any real velocity.
  • the mass and damping are dynamically
  • the updated pointer 210 location is then sent by the CCS 102 to each of the computing devices 104 and is used by the CIA running on each of said computing devices 104 to update the displayed location of the pointer 210.
  • the plurality of users can watch the pointer 210 move, not based on their own individual input, but based on the overall collective intent of the group.
  • the group intent vector can be computed from the plurality of user intent vectors as a simple average, or may be computed as a weighted average in which some users have more influence on the resulting collec ive group intent than other users.
  • the weighting of each user can be derived based on user scores and/or user synchronic!ty values (also referred to as synchrony values) earned during prior interactions with the system 100.
  • each user may be assigned one or more variables that represents how his or her input should be weighted with respect to othe - users in the swarm.
  • the variable is called the user contribution index and is updated
  • constructive input i.e. input that is supportive of the collective intent
  • destructive input i.e. input that is substantially resistant to the collective intent
  • Those users who are supportive to the emerging consensus are determined computationally by the CCS 102 by repeatedly comparing each user' s user intent vector with the group intent vector. The more aligned that user' s user intent vector is with the direction of the group intent vector, the more collaborative that user is behaving . The further the user intent vector is from the direction of the group intent vector, the less
  • the synchronicity value may be an instant synchronicity value, i.e. one at a certain instant, in time, or may be a session synchronicity value representing the overall user synchronicity for one or more sessions.
  • the synchronic! ty value for each individual user is determined by the CCS 102 by repeatedly comparing the user intent received from each computing device 104 (representing the user input reflecting the user's intent to move the graphical obj ect of the pointer 210 in a given direction) with the group intent, derived from all user intents .
  • the synchronicity value of the individual user is determined but computing the difference between the user intent and the group intent.
  • the synchronicity value may be an instant value, i.e., based on a single comparison of the user intent to the group intent at one point in time, or may be synchronicity value over a specific period of time, e.g. an average of the
  • the user synchronicitv value each individual user represents at least in part that user's contribution to the
  • each individual's synchrony value ranges between an upper bound value and a lower bound value.
  • the synchITO ⁇ 1city value ranges between +1 to -1, with the value +1 (the upper bound) being assigned when the user intent vector is substantially aligned with the group intent vector, and with the value of -1 (the lower bound) being assigned when the user intent vector is substantially in the opposite direction of the group intent vector, with all values between +1 and -1 being used to represent varying degrees of alignment. For example, if the user intent vector is 90 degrees out phase with the group intent vector, a value of 0 is assigned, for that is halfway between fully convergent and fully divergent.
  • a skilled user is one who is able to convey his individual intent as input, but do so in a cooperative manner .
  • Such a user will maintain a positive synchrony value during much of the session, for he or she is being supportive of the group intent .
  • a user who maintains a positive value will be awarded more points and be assigned a higher user contribution index than a user who does not.
  • the user' s synchronicitv values are computed as a percentage from 0% to 100%, for that is often an easier metric for users to understand.
  • the session synchronicity value of 100% means the user has been perfectly in sync with the swarm.
  • the session synchronicity value of 0% means the user has been
  • Session synchronicity values between 0% and 100% reflect relative synchronization with the swarm, with a 50% synchronicity value meaning the use was neutral with respect to the swarm. This is described in more detail later in this document.
  • an average or mean
  • synchronicity value is computed for the user over some number of prior questions. For example a "sync 5" synchronicity value can be computed as that user' s ave age synchronicity value over the last five sessions. This is a highly useful value for it indicates how cooperative the user has been over a recent period of time.
  • the "sync 5" synchronicity value can be used in combination with other time-histories, such as a
  • “sync 50" synchronicity value which indicates the average synchronicity value for that user over the last 50 sessions, in order to compute that user's weighting value in the swarm.
  • synchronicity value may be time-weighted such that time steps near the end of the session time period are more heavily -weighted than time steps near the start of the time period.
  • the CCS 102 determines at least one user assessment based at least in part upon one of more user synchronicity values. For examples, one
  • assessment may be configured to determine whether the user is categori zed as " f1exib1e” or "entrenched” . In another example, one assessment may be configured to determine whether the user is "constructive" or
  • FIG. 4 shown is an exemplary display interface 400 as displayed on the computing device 104 being used by one user of a group, shown at. a moment in time after the group has successfully
  • the pointer 210 on one of the input choices 208, selecting the input choice 208 as the target, thereby collaboratively choosing the answer. Shown are the group name 204, the target area 206, the plurality of input choices 208, the communication menu 212, the chat 'window 218, the chat input box 220, the current member list 222, the statistics display 224, the invite button 226, a prefix text 402, a target text 404, a group cohesiveness score indication 406, a session
  • synchronic!ty value score indication 408 a points indication 410, an answer window 412, an answer options tab 414, a replay swarm icon 416, and a Tweet answer icon 418.
  • the target is "Gremlins"
  • the target is graphically displayed to each user on the screen of his or her computing device 104 (as controlled by the CIA software running on that device 104) .
  • the graphical display includes the answer window 412 including the prefix text 402 "UNUM says:” along with the chosen target : “Gremlins” .
  • the answer is also displayed in the chat window 218, as if communicated by the
  • the group cohesi eness score indication 06 reflecting the synchro icity of the group, is shown of 84%, 'which indicates that the group was 84% aligned in their imparted motion of the pointer 210.
  • cohesiveness score indication 406 includes the text "GROUP SYNC:”
  • the group cohesiveness score of 84% shows strong convergence of group members, reflecting that, the swarm intelligence spoke with high “conviction” when answering this question.
  • a low group cohesiveness score ' would reflect a low conviction for the swarm
  • the group cohesiveness score may be repeatedly reported to and repeatedly displayed by each of the computing devices 104, for example during the session.
  • the group cohesiveness score may have an upper bound and a lower bound, wherein a group cohesiveness score at the upper bound indicates that the plurality of real-time user intents are substantially aligned with each other, and a group cohesiveness score at the lower bound indicates that the plurality of real-time user intent values are substantially misaligned with each other.
  • the lower bound is essentially 0, as the summation of the user intent vectors, being opposite (exactly misaligned) , cancel each other out.
  • the CCS 102 determines at least one group assessment based at least in part upon one of more group cones iveness scores. For examples, one
  • assessment may be configured to determine whether the group is categorized as "flexible” or "entrenched”.
  • the group cohesiveness score may be repeatedly calculated by the CCS 102 during the session and
  • the real-time user intent values are determined to be substantially aligned with each other (i.e. at or near the upper bound) when their vector directions are substantially the same in at least a plane.
  • the real-time user intent values are determined to be substantially misaligned with each other (i.e. at or near the lower bound) when a summation of their vector directions substantially cancel each other out, resulting in a near zero resultant.
  • the session user synchronic!ty value score indication 408 is a statistical indication of how well the particular user of this computing device 104 was aligned in his input with the swarm as a whole.
  • the session synchronicity value score indication 408 includes the text "YOUR SYNC:" and value of 91%. In this case, the user was very highly aligned, achieving a 91% synchronicity value.
  • the points indication 410 is also displayed in the answer window 412, indicating the number of points earned by this user as a result of his or her
  • the points indication 410 also includes the text "POINTS:"
  • the answer options tab 414 which gives users options related to the answer that was just reached by the swarm.
  • the user can selectively Tweet ® the answer by selecting the Tweet answer icon 418. This triggers a routine within the CIA that, sends a Tweet request to the CCS 102 software, which then sends an automated Tweet to Twitter.
  • the Tweet includes the question and the selected answer.
  • the Tweet also
  • the Tweet includes a numerical indication of the number of users who participated in answering the given question, thus conveying the size of the swarm intelligence which produced this Tweet.
  • the Tweet also includes a hashtag, for example "#UNUMsays", as well as an indication of the group cohesiveness score. In this way, the swarm
  • the decision to Tweet an answer is posed by the software to the swarm.
  • invention described herein enables the formation of a swarm intelligence, enables that swarm intelligence to answer questions, enables that swarm intelligence to consider the answer that emerges and decide if that swarm intelligence wants to Tweet the answer publically.
  • each individual user can select a replay swarm icon 416.
  • the session resulting in the cu rent answe is replayed on the display.
  • the session replay is unique in that it displays an indication of the input of all users in the group at the same time (i.e. the swarm input), giving insight into how the swarm converged upon the collective answer.
  • the video of the swarm input is displayed in high speed
  • FIG. 5 a frame of an exemplary session, replay video 500 is shown . Shown are the target area 206, the plurality of input choices 208, the
  • the session replay includes the question asked, the input choices 208, and the graphical indication of the trajectory taken by the pointer 210 during the answer period. Also displayed is the graphical indication of the input provided by each user of the swarm at each time-step during the answer session.
  • the graphical magnet icon 306 is displayed for each user, the size and orientation of each magnet icon 306 with respect to the pointer 210 indicating the magnitude and direction of that user' s user intent vector (magnitude and direction) upon the pointer 210 at each given moment in time.
  • 8 users were participating in the swarm, collaboratively moving the pointer 210 to an answer. This method is scalable to much larger numbers of users.
  • the software running on the local user's computing device 104 can be configured to show all magnet icons 306 in the replay as a uniform, color except for the magnet icon 306 representing the time-history of that particular user's input. For that user, the magnet icon. 306 can be shown as an. alternate color with visual contrast .
  • the user can observe the swarm of many magnet icons 306 as the history of the session is replayed and identify his or her own magnet icon among the swarm of many magnet icons 306 because his own magnet icon 306 is displayed in the alternate color.
  • the local software on each computing device 104 is configured to identify which magnet icon 306 in the replay is associated with the user of that computing device 104. Such identification can be achieved by associating each magnet icon 306 in the replay with a unique user ID value stored in memory.
  • the present invention employs a number of inventive systems and/or methods for
  • each group is a collection of intelligent members (users) that are networked together in real-time, each of them providing col laborative input that' s
  • a first approach is to dynamically modify the swarm population by purging the swarm of one or more of its currently low-performing members (the input from said members determined to be substantially out of sync 'with collaborative 'will of the swarm, i.e. having a low synchronicity value) and/or setting a statistical
  • a second approach is to dynamically modify the connection strengths within a given group population by adjusting the weighting assigned to the inputs from each individual user, the weightings assigned to each given user being modulated to improve overall group
  • the CCS 102 software is selectively configured to increase the 'weighting of inputs from high-performing members of the group in terms of their collaborative behavior, and decrease the
  • the CCS 102 In order for the CCS 102 to purge users from the group, institute thresholds that limit entry into the group, and/or dynamically modify the connection strengths within the group, the CCS 102 must quantify swarm performance as well as user performance in the context of collaboration, for determining levels of collaborative performance is used as the basis for dynamic modulation of the group. To perform such quantification, the group cohesiveness score (representing the group synchrony) and the user synchronic! ty value (synchrony value) is used.
  • synchrony is determined computationally by the software running on the CCS 102 based on the degree of alignment (in direction and magnitude) among the user input collected from all member of a swarm, during a response. Because the degree of alignment changes at every time-step, the software running on the CCS 102 is configured to integrate over the response period, producing time-weighted average . In this way, the synchrony computed during a single question/answer session is the time-weighted average of the instantaneous synchrony (i.e. alignment among input vectors) across all time steps.
  • the two types of synchrony are computed by the CCS 102 software and communicated to each of the peers : group synchrony and individual synchrony. These are described in detail as follows :
  • the group cohesiveness score is an indication of the collaborative coordination of the group as it answers a question or completes a task, derived by computing the degree of alignment among the full set of user intent vectors from, all participating users in the group, integrated across all time steps of the session. In many current embodiments, this value is expressed as a
  • the computation is configured such that if, in theory, all of the users of a group coo dinate perfectly during the session (i.e. all users impart input vectors of the exact same magnitude and direction at every time step across the session) , that group would deemed to have a group cohesiveness score of 100%. In practice, this rarely happens.
  • the outcome of the session is one where the central tendency of the group leads to a coherent answer through the motion of the pointer 210. This generally translates into a group cohesiveness score between 65% and 90% . Conversely, if all members of the group are pulling in the exact opposite directions (i.e. all user intent vectors perfectly cancel out), the pointer 210 will not move at all, resulting in a
  • the inventive system still identifies unproductive swarms where the pointer 210 sputters, moving in one direction and another, but never finds enough consensus to drive the pointer 210 to the answer. Such sessions generally have the group cohesiveness score of between 10% and 35%.
  • the CCS 102 software is configured to measure and report the group cohesiveness score to every user after every session, (i.e. every collaborative answer ⁇ .
  • the group cohesiveness score is measured and report to every user after every session, (i.e. every collaborative answer ⁇ .
  • users By giving users a direct, and easy to understand measure of the collaborative coherence of the group, they can understand if the group is performing well together and adapt their actions accordingly.
  • points or credits
  • the points are scaled by group cohesiveness score.
  • all users are rewarded when the group shows high synchrony (i.e.
  • the user synchronicity value is a. time-weighted average that's integrated across all time steps, but. in this case the synchronicity val e is a measurement of how well aligned a single user is with respect to the group as a whole. Because the synchronicity value is personalized for each user, the CCS 102 software must compute the user synchronicity value independent1y for each member in the group, indicating how well aligned that user' s input vector was with the overall group input vector. The user with the high synchronicity value (>65%) during the session is deemed to have been highly supportive of the resulting consensus, contributing to the emergent response .
  • the user with the low synchronicity value ( ⁇ 35%) during the session is deemed by the software to be obstructionist, standing in the way of compromise and consensus .
  • the CCS 102 software measures and reports each user synchronicity value after each session sending each user their personal user synchronicity value for display on their own computing device 104,
  • points (or credits) are awarded to the user
  • the number of credits or points is based at least in part on that user' s user synchronicity value and/or the group cohesiveness score.
  • user points (or credits) are awarded based 60% on that user's user synchronicity value and 40% on the overall group cohesi eness score. In this way, users are incentivized to perform collaboratively as individuals, while also being incentivized to push the swarm, to behave collaboratively overall. This is highly effective.
  • an exemplary display interface 600 is shown during a session determining whether to eject a specific member from the group. Shown are the prompt bar 202, the group name 204, the target area 206, the plurality of input choices 208, the pointer 210, the commun cation menu 212, the chat window 218, the chat input box 220, the current member list 222, the statistics display 224, the invite button 226, the question display 302, the countdown clock 304, the magnet icon 306, and the flag icon 602, Each group is configured to be able to eject, or purge members of the group who consistently show low user synchronicity values over a certain number of sessions.
  • the determination is based on a user' s average user synchronicity value over the last 5 sessions (referred to herein as that, user's "Sync 5" synchronicity value)
  • “banishment decisions” are posed to the group itself, • which uses collaborative motion of the pointer 210 to decide if an identified low-performing member should be banned for low performance.
  • the benefit, of using the Sync 5 user synch onicity value is that users are not punished for a single divergent answer, or even a few divergent answers, but a string of them. This helps to differentiate between users who just disagree with a single question versus users who are deliberately being obstructionist to the swarm's overall performance.
  • the Sync 50 is also computed, which is the time average of the user' s user synchronicity value over the last 50 session. This value is used in
  • Sync_5 user synchronicity value and the Sync_50 user synchronicity value are effective, but obviously values averaged over a different number of sessions could be used by the CCS 102 software.
  • the key is for the
  • the present invention enables the user to create a new group by giving the new group a name, assigning it a theme, and including a description of the new group's intent and/or philosophy.
  • the user creating the new group can assign an entry threshold value that indicates a level of historic user
  • the Sync 50 user synchronicity value is used.
  • the group creator might indicate that only users with a Sync 50 greater than 35% can enter the new group. This ensures that deliberately obstructionist users (based on historical performance) can't enter.
  • the system of the present invention enables a virtual lobby interface 800 included in the display interface, the virtual lobby interface 800 indicating a plurality of distinct groups for users to join, each of the plurality of groups having a different entry threshold, or
  • Shown in FIG. 6 is an exemplary user display interface that supports the purging methodology described herein.
  • the CCS 102 has identified that a meraber of the group has been assigned a Sync 5 user synchronicity value below the pre-assigned threshold (for example, a user synchronicity value below 20%) .
  • the CCS 102 software sends an automated question to all members of the group, asking if the low-performing member should be purged from, the group .
  • the question automatically posed to the group by the CCS 102 includes the unique user name of the low performing member ("JaneDoe" in the exemplary session) and an indication of the threshold that was fallen below ("Sync 5 ⁇ 20%” in the exemplary session) .
  • the members of the group then engage in the collaborative session, providing input in real-time that is numerically combined into the group intent.
  • the CCS 102 software automatically sent each member of the swarm a target area including the input choices 208.
  • the set of six input choices 208 includes:
  • CCS 102 software is configured to monitor the future user synchronicity values for that user, giving that user a defined amount of time (or defined number of session) to raise his user synchronicity value above the defined threshold. For example, the user may be required to get his Sync 5 user synchronicity value above 35% within the next ten sessions, or ejection of that user will
  • one user of the group can initiate a purge session by clicking on a particular user's username (as shown in the list of current members) and selecting a "purge user" option from the board selection drop-down menu 214. In preferred embodiments, this can only be done if the user synchronicity value or other measure of performance of the user to be purged has fallen below the threshold value. In some such
  • the flag icon 602 appears in the list of current members next to the usernames of users whose user synchronicity value fell below said threshold, thus alerting the other members of the low performance, and alerting the other users that such "red flagged" users can be selected for possible purge question put to the group.
  • the CCS 102 does a periodic purge that does not identify the specific username of the potentially purged user when posing the question, to the group.
  • the CCS 102 automatically sends the question - "Should we purge the lowest performing member of the group?" The group must now respond.
  • the dynamic is interesting because members of the group do not know if they are the lowest performing member.
  • such purge sessions are triggered at regular time intervals.
  • such purge sessions are triggered when the group cohesiveness score falls below a threshold. This is highly effective because the group cohesiveness score is a representation of how collaboratively
  • the CCS 102 can be configured to ask "Should we PURGE the lowest performing 10% of our members?" This enables the swarm to purge many members at once if they are not performing well. Again, the dynamic is quite interesting and engaging for users because they don' t know if they are among the lowest 10% that will get purged. In this way, the swarm can self-moderate itself, enhancing its own configuration for optimal performance, with assistance from, the automated agent of the CCS 102 soft'ware .
  • an exemplary display interface 700 is shown during a session determining whether to allow a specific member to join the group. Shown are the prompt bar 202, the group name 204, the target area 206, the plurality of input choices 208, the pointer 210, the communication menu 212, the chat window 218, the chat input box 220, the current member list 222, the statistics display 224, the invite button 226, the question display 302, the countdown clock 304, and the magnet icon 306.
  • the collaborative group is not only empowered to make
  • the swarm can be configured when created to be " dmit only” in which case, users must be collaboratively granted access.
  • This designation (or similar designation) is displayed in the system lobby display. If the swam is identified in the lobby display as "swarm admit only", the user may not immediately join the group, but the user may select a displayed button marked "knock". When a user knocks on a swarm (i.e. selects the knock button, whereby an
  • the CCS 102 software is alerted that the user wants to enter that particular swarm and because that swarm is listed in the CCS 102 database as being "swarm admit only", the CCS 102 software executes a routine that puts the admission question to the group.
  • the swarm intelligence can then collaboratively decide if it wants to allow the given user to join, or reject the request for admission.
  • the CCS 102 has received an indication that the user BIG DAVE has indicated that he or she wishes to enter the group "Swarm 001" . Further, responsively the CCS 102 determined that the group
  • the CCS 102 determines an ordered rank of a plurality of users based at least in part upon at least one synchronicity value associated with each of the plurality of users. In some embodiments the CCS 102 determines an ordered rank of a plurality of groups based at least in part upon at least one group cohesiveness score associated with each of the groups
  • the CCS 102 might provide a link to further stats or information about that user, possibly including a link to his or her Facebook ® page or Twitter ® handle. In this way, the members of the swarm can assess who this user is, and how collaborative this user has been during his prior participation within the system.
  • the current members of the swarm SWARM 001 then engage in the collaborative control process, providing input in real-time that is numerically combined into a singular intent of the swarm intelligence, as shown by FIG . 7,
  • the set of six input choices 208 includes: “no”, “yes”, “not now”, “yes”, “no” and, "bad question”.
  • the users then collaboratively engage, enabling the swarm, intelligence to converge on the target input choice 208. If the target is "yes”, the identified user is granted entry into the swarm. If the target is "no” the identified user is not granted entry into the swarm. If the target is "not now” the identified user is informed by the CCS 102 software: “maybe... try again later. " In this way, the collaborative swarm
  • the swarm can be any suitable swarm.
  • the swarm can be any suitable swarm.
  • the group configuration configured to dynamically adjust the group configuration, not only by selectively ejecting users from, the swarm and/or admitting members to the swarm, but by adjusting the relative weighting of the input received from current members of the swarm. More specifically, in some
  • dynamic algorithms are used to increase the weighting that, certai users have upon the collective pull of the pointer 210, while decreasing the weighting that other users have upon the collective pull of the pointer 210.
  • the CCS 102 can be configured to compute and store a weighting value for each user, based on that user's historic user synchronicity values. Users who show a time history of high user synchronicity values are assigned a positive weighting value, while users 'who show a time history of low user synchronicity values are assigned a negative weighting value. These weighting values are updated regularly by the CCS 102, ideally after each session that a user participates in, because the user' s performance during that session likely
  • the swarm intelligence is adapted over time, strengthening the connections (i.e. input weighting) with respect to the more collaborative users in the swarm, and weakening the connections with respect to the less collaborative users in the swarm.
  • the collaborative swarm is dynamically adjusted in an
  • the CCS 102 computes the Sync 5 user synchronicity value and Sync 50 user
  • Sync 50 user synchronicity value for each user, based on the user' s performance during multiple sessions . For example, the user might have participated in 50 sessions as a member of multiple groups. Thus the Sync 50 user synchronicity value that is stored and updated on the CCS 102 (and related database) is swarm-independent .
  • the CCS 102 computes the weighting value for that user based on his Sync 5 user synchronic! ty value and Sync 50 user synchronicity value (reflecting the user's user synchronic! ty value over the last 5 and last 50 questions respectively) .
  • the weighting value is computed as follows :
  • This equation assigns a 'weighting value that's 40°; dependent upon the user' s S /nc 50 user synchronicity value and 60% dependent upon the user's Sync 5 user synchronicity value, thereby giving greater importance to the user' s more recent behavior, but still considering the longer term behavior of that user.
  • this equation is structured mathematically such that users who earn user synchronicity values at or near a neutral performance level of 50% have no change in weighting, and users who have user synchronicity values much higher than the neutral value of 50% have a higher weighting, this higher weighting value topping out at +10%. Users with user synchronicity values substantially below 50% are computed to have a negative weighting value that maxes out at -10%.
  • weighting values could be defined with a larger range, for example -20% to
  • the present invention includes one or more user-selectable mode when asking a question that also changes the dynamics of the collaborative answer.
  • the present invention includes a user selectable mode called "gravity mode" that is accessible from the physics selection drop-down menu 216. The gravity mode is engaged during the session such that the pointer 210 experiences the restoring force that pulls the pointer 210 back to the point
  • the system can be configured to require that at least 80% (i.e. 80 users of the 100 in the group at the present time) are pulling in a substantially similar direction to overcome gravity and position the pointer 210 on the desired target. This mode thus enables a high barrier for collaborative decision making, requiring the group to have more "conviction" in the resu11ing response .
  • the level of gravitational force is user-selectable, thereby ad usting the level of conviction required to overcome gravity and reach the target answer .
  • an example display interface of the virtual lobby interface 800 is shown. Shown are a group directory 802, the plurality of group names 204, a plurality of group themes 806, a plurality of group cohesiveness score indications 406, a plurality of information icons 812, a plurality of statistics icons 814, a plurality of log icons 816, a plurality of
  • favorites icons 818 a number of users in the group 820, a plurality of maximum number of users 822, a plurality of unlocked icons 824, a locked icon 826, a favorites section 828, a swarm creation section 830, a plurality of user input areas 832, a make private selection box 834, and a create button 836.
  • the virtual lobby interface 800 is accessible to computer users on computing devices 104 either through the CIA running on their computing device 104, or through a standard web browser (if the virtual lobby interface 800 is created as a standard html webpage) .
  • the virtual lobby interface 800 includes the group directory 802 of available groups that users can join and then participate in real-time collaborative intelligence processes .
  • the virtual lobby interface 800 is not real-time, but employs more traditional methods known to the art when joining chat rooms.
  • the virtual lobby interface 800 is divided into a number of sections .
  • One section is the group directory 802 labeled as "UNU Centra1" .
  • users can browse the available groups, each of said groups being associated with a theme that governs the type of
  • the group directory 802 in the embodiment shown comprises a table, with a row for each group included in the directory. Tnform.at.ion included in the row for each group includes the group name 204, the group theme 806, the current number of users in the group 820, the maximum number of users 822, and the current group cohesiveness score.
  • the group theme 806 is a general description of the area of focus for the group, for example, investing, music, politics or technology.
  • the group cohesiveness score is low, users may not want to enter that swarm because it. means the group is not being highly collaborative. The low group
  • Also included in the row for each group is a plurality of tool icons. Included in the tool icons of the exemplary lobby interface 800 of FIG. 8 are the information icon 812, the statistics icon 814, the log icon 816, and the group member icon. When the user selects the information icon 812 for one group, a display of additional information about that swarm is shown. When the user selects the statistics icon 814, a display of statistics of the group is shown. Statistics may include a number of questions asked by the group during one or more periods of time, an average number of users that participated in the group during one or more periods of time, and the average group cohesiveness of the group during one or more periods of time.
  • the average group cohesiveness may be determined by finding the mean of a series of repeated group cohesiveness scores over a specific period of time.
  • the mean is time-weighted such that time-steps near the end of the time period are more heavily weighted than time steps near the start of the time period.
  • the period of time may comprise a plurality of completed question-and-answer sessions.
  • the log display has been disclosed in the related applications .
  • the log display may
  • the CCS 102 archives not just a history of questions and answers for each swarm, but archives the replay data associated with each of said questions and answers .
  • the replay data includes locative data for the pointer 210 and each of the magnet icons 306, said data stored at regular time intervals over the period of a response to a question. For example, pointer location coordinates along with magnet icon 306
  • positions, orientations, and size data may be stored every 0.25 seconds during the period of the response to the question.
  • data related to the pointer 210 being over input choices 208 may also be stored.
  • magnet icon data is stored relative to pointer 210 location, for example as a distance vector from the center of the pointer 210, the distance vector having a size and orientation relative to the center of the pointer 210.
  • the favorites icon 818 indicates which of the groups are included in a "favorites" list.
  • the favorites list includes groups that user has selected as favorites, groups that have been created by the user, and private swarms that the user has been invited to.
  • the groups includes in the user's fsLVorit.es are X-Men, Bigbrain, HumanZoo, OuterLimits, and 3D-Makers groups, as indicated by the highlighted (white) star icon. Groups not included in the user's favorites list are indicated by the unhighlighted (black) star icon .
  • Some groups displayed in the group directory 802 are configured to have limitations to group membership, as irevious 1y described. hese groups are indicated by either the locked icon 826 or the unlocked icon 824 next to the group name 204. In the group directory 802 portion shown, the HumanZoo and 3D-Makers groups include the unlocked icon 824, indicating that it is currently possible to join those group's if the membership
  • the Seance group includes the locked icon 826, indicating that it is not possible to join that group at this time.
  • the locked icon 826 may be displayed for one of a plurality of reasons, for example - the swarm may be locked because it is private and requires an invitation or password to be joined by the user.
  • the swarm may be locked because it has an entry threshold such that users must have scores and/or statistics related to thei historical performance that are above the entry threshold to be granted access.
  • the swarm may be locked because the swarm is configured to require group approval for new users joining.
  • the swarm may be locked because it has reached its real-time group size limit and thus cannot accept any additional users at the present time.
  • the swarm creation section 830 allows users to create their own swarm.
  • the user can define the name of a new swarm, give the new swarm, a. theme, and optionally make the new swarm a private swarm that requires a password, by selecting the make private selection box 834.
  • users are further given the ability to invite their friends to the new swarm by accessing their Facebook ® friends and/or Twitter ® followers .
  • the favorites section 828 of the display interface allows users to track swarms that are of particular interest to them.
  • the favorites section. 828 comprises a table including the swarms included in the user's favorites list.
  • the favorites section 828 is formatted similarly to the group directory 802 table, including the UNUM name, theme, number of users 820, maximum, number of users 822, and icons 812, 814, 816, 818 for each swarm included in the favorites section 828.
  • the favorites section 828 may also include the locked icon 826 or the unlocked icon 824 for the group, as applicable.
  • the present invention allows users to enter swarms, exit swarms, and create swarms.
  • the historical performance for users (for example their score, credits, ranking, rating, and synchronic!ty values) are maintained by the CCS 102 for participation across all swarms .
  • a user can earn points by participating in a variety of swarms, public and private, although they can only be in one swarm at a time. That's because swarms require real-time participation.
  • users are given the ability to configure new swarms by setting parameters that indicate: (a) whether the new swarm is private or public, (b) whether the new swarm supports adaptive weighting or all users should always have equal weighting, (c) whether the swarm supports automated purging or the purging of users should always be user initiated, (d) whether the swarm is supports "swarm admit only” or anyone can join the swarm without the swarm intellect making an assessment, (e ) whether the swarm supports an entry threshold and if so, what level it should be, (f) whether the swarm supports an ejection threshold and if so, what the level should be.
  • each swarm can be linked to one or more official Twitter ® accounts, for the sending of Tweets that represent the official voice of that swarm intelligence.
  • a set of six potential answers is displayed to a plurality of etworked users along with a textual prompt.
  • the answer options and prompt are
  • the plurality of networked users By engaging a user interface associated with each of the separate local computers, the plurality of networked users 'work together as a unified dynamic system, collectively moving the graphical puck 210 (also referred to as the pointer) from one displayed starting location to one displayed target location associated with one of the six input choices 208 (also referred to as answer choices) . In this way, the group of networked users make a real-time decision as a collective
  • the decision period the time period from just before the puck 210 starts moving until the puck 210 lands upon (and selects) one of the provided answer choices 208 is referred to herein as the decision period.
  • each of the plurality of users in the collective intelligence influences the motion by imparting his or her own personal intent regarding the motion of the puck 210 at repeated moments in time (i.e. continuous time steps) .
  • users do this by manipulating the graphical magnet icon 306, the location and orientation of the magnet with respect to the puck 210 defining the
  • the present invention allows a plurality of networked users to form a system that makes decision in much the same way that a biological brain makes
  • the decision-making mechanism in a biological nervous system is described as "a competition between mutually interacting populations of excitable units (i.e. neurons) that, accumulate noisy evidence for alternatives and 'when one population exceeds a threshold level of activity, the corresponding alterative is chosen"
  • the present invention allows this same decision-making process to occur, not by connecting a plurality of neurons into a brain, but by connecting plurality of brains into a larger structure referred to herein as a "hyper-brain". And just like a brain
  • the present invention enables a system of networked users to combine their noisy and disparate input in real-time, producing a rapid and definitive decision. Once a decision is made by the system, of users, it is desirable to analyze how that, decision was arrived, at from the noisy collection of inputs. More
  • Stalemates occur when the factions pulling towards differe t optio s impart force vectors that balance to a net force of zero or nearly zero.
  • Such stalemates are only resolved if one or more of the users changes his or her pull from, one option to another. In other words, deadlocks are resolved if one or more users defects from one faction to another.
  • the system may include a large number of users working together to move the puck 210, it's often the case that, many users change factions at any moment in time during the decision period. This enables a complex negotiation among all the participating users, everyone pushing and pulling, trying to find a solution that best satisfies the collective will of the unified system.
  • the present invention provides such analysis.
  • the present invention also enables faction assessments to be visualized. While the present invention is described herein by embodiments in which one option being selected from a set of six options (input choices 208), these methods can be
  • the methods and systems enable the plurality of networked users to participate in a real-time process in which the question or other textual prompt is
  • the local (portable) computing devices 104 are in
  • the system and methods of the present invention enable the plurality of networked users to respond to the prompt as a unified dynamic system, collectively selecting one response from the set of possible responses. In many embodiments, the users do this through real-time closed- loop control of the
  • the collaborative pointer 210 in which the plurality of users work in synchrony to move the pointer 210 from a starting location to a location associated with the selected response.
  • the users impart their individual intent with respect to the motion of the collaboratively controlled pointer 210 by
  • the CCS 102 receives the plurality of user intent vectors and
  • the present invention enables the plurality of users to work together as the real-time closed-loop collaborative intelligence that expresses a singular group intent that can answer questions, make decisions, or otherwise provide collective responses to a textual prompt.
  • the methods intervening software and hardware to moderate the process, closing the loop around the disparate input from each of the many individual participants and the singular output of the group.
  • each individual user (“participant") engages the user interface on the computing device 104, conveying his or her individual real-time intent with respect to the motion, of the collaboratively controlled pointer 210, while simultaneously watching the real-time motion resulting from the group intent. This closes the loop around each user, for he is conveying individual intent while also reacting to the group' s emerging will.
  • a time period from the start of a question period (e.g. when the word "GO" appears on the plurality of computers) to when the target is selected is referred
  • the group of users works as a real-time dynamic system, to move the puck 210 from the staring location of the selection.
  • the unique analysis is performed by the Central Collaboration Server 102
  • a "faction" is a sub-group of the plurality of users working as a unified dynamic system to answer a. question, by moving the collectively controlled pointer 210 from, a starting location to a location associated with an answer (i.e. selecting a target) .
  • the faction is a sub-group at a current moment in time who are all conveying User Intent Vector forces that aim. to move the collectively
  • the faction of users is defined herein as a. sub-group of the total plurality who are applying individual user force vectors which, aim towards the same choice of the plurality of available answer choices 208. If, for example, there are six available answer choices 208 (as shown around the hexagon of FIG. 2), then at any moment in time during the
  • the present invention stores a representation of each of said factions in the memory of the CCS 102, the representation in memory indicating which user of the plurality of users is associated with each of the factions at various moments in time during the decision period.
  • the CCS 102 stores in memory a list of user identifiers for each of six factions for each of a plurality of discrete time-steps during the decision period.
  • the time step is a quarter second.
  • the CCS 102 stores the list of user identifiers in memory indicating which user is currently part of the six factions (i.e. pulling towards each of the six choice solutions) .
  • the CCS 102 stores in memory a "null faction" which includes a list of user identifiers for those users who are not currently pulling towards any of the six choices 208.
  • Some of those users may be classified as "dise gaged" for that time step, meaning they are not currently pulling on the puck 210 in any direction. This is likely because their magnet icon is not within proximity of the collectively controlled puck 210. Or it may be because the user is pulling on the puck 210 (via the magnet) in a direction that is not associated with any of the six factions, instead falling in a direction that falls between the direction associated with each faction. This will be made more clear with respect to the additional figures below .
  • an exemplary target area is shown during three different time steps during an exemplary decision period, A first time step target area 900 is shown at s time step of 0 seconds elapsed since the start of the decision period. A second time step
  • a target area 1000 is shown at a time step of 6 seconds elapsed.
  • a third time step target area 1100 is shown at a time step of 12 seconds elapsed. Shown are the pointer 210, the plurality of magnet icons 306, the plurality of answer choices 208, a first selection target 902, a second selection target 904, a third selection target 906, a fourth selection target 908, a fifth selection target 910, a sixth selection target 912, a first faction 914, a second faction 916, a third faction 918, a fourth faction 920, a fifth faction 922, and a sixth faction 924.
  • Each selection target 902, 904, 906, 908, 910, 912 represents a target location wherein the pointer 210 is moved under collective control to one selection target 902, 904, 906, 908, 910, 912 to select the corresponding input choice 208,
  • Each faction 914, 916, 918, 920, 922, 924 corresponds to the same-numbered selection target 902, 904, 906, 908, 910, 912, i.e. the first selection target 902 corresponds to the first faction 914, etc.
  • FIGS. 9-11 show the three different t ime step target areas 900, 1000, 1100 depicting three different time step "snapshots" during the exemplary decision period in which a group of users are working as a unified dynamic system to answer a question as a collective intelligence.
  • the question has just been asked, the word "GO" being displayed to the plurality of users on each of their computing devices 104.
  • decision period is 0 seconds. At this moment in time, all of the users who are participating are applying their initial pull on the puck 210 by positioning their
  • the first time step target area 900 represents what the CCS 102 stores in memory, which is the location of the puck 210 and the location and orientation of each of the plurality of magnet icons 306 (representing each user intent vector) , each magnet icon 306 controlled by one separate user on one separate computing device 104. Also shown is the layout of the answer choices 208 and selection targets 902, 904, 906, 908, 910, 912 around the hexagon shape. In this example the answer choices 208 are represented as numbers “1", “2", “3”, “4", "5", and “6”. It is understood that these answer choices 208 are generally words or phrases such as "Bill Clinton” and "George Bush", if the question
  • the graphical selection targets 902, 904, 906, 908, 910, 912 are also shown.
  • the first time step target area 900 As shown in the first time step target area 900, at zero seconds elapsed, i.e. time step ⁇ 0 seconds, all of the participating users are ready to pull on the puck 210 in different directions. Many of those users intend to pull the puck 210 towards one of the six selection targets 902, 904, 906, 908, 910, 912 around the hexagon, as shown by the location of the plurality of magnet icons 306 shown, in FIG. 9.
  • What, is needed, however, is a rapid analysis method to determine which users are pulling towards 'which answer choice 208 at each time-step during the decision period. What is also needed is a way to group users into "factions" such that factions can be tracked over time as they form and dissolve, their populations of users changing during the decision period.
  • one magnet icon 306 is included in the first faction 914, three magnet icons 306 are included in the second faction 916, one magnet icon 306 is included in the third faction 918, two magnet icons 306 are included in the fourth faction 920, one magnet icon 306 is included in the fifth faction 922, and one magnet icon 306 is included in the sixth faction 924.
  • the second time step target area 1000 shows the decision period at a later time step, wherein six seconds have elapsed since the start of the decision period.
  • the combined group vectors have shifted the collectively controlled pointer 210 closer to the fourth selection target 908.
  • the magnet icons 306 have shifted, with seven magnet icons 306 (each magnet icon 306 representing one user) pulling towards the fourth selection target 908 and comprising the fourth faction 920, three magnet icons 306 pulling towards the second selection target 904 and comprising the second faction 916, two magnet icons 306 pulling towards the fifth selection target 910 and comprising the fifth faction 922, and one magnet icon 306 pulling towards the first selection target 902 and comprising the first faction 914.
  • No magnet icons 306 are pulling towards (or approximately towards) the third selection target 906 and the sixth selection target 912, whereby there is no third faction 918 or sixth faction 924 for the current time step.
  • the factions are consolidating, with the fourth faction 920 gaining support and no third faction 918 or sixth faction 924.
  • the decision period is shown at an even later time step, wherein twelve seconds has elapsed since the start of the decision period.
  • the combined group vectors have shifted the pointer 210 even closer to the fourth selection target 908, and the fourth selection target 908 is very close to being selected as the target.
  • the magnet icons 306 have shifted again as the users change their input during the decision period. Even though nine magnet icons 306 appear to be pulling towards or approximately towards the fourth selection target 908, only four magnet icons 306 are included within the fourth faction 920.
  • Two magnet icons 306 are included in the fifth faction 922, and one icon is included in the first faction 914.
  • the second faction 916, the third faction 918, and the sixth faction 924 have no magnet ico s 306 are therefore not shown .
  • the grouping of magnet icons 306 in factions related to the selection targets 902, 904, 906, 908, 910, 912 changes over time.
  • the criteria for determining which magnet, icons 306 belong to a given faction is important, as seen in FIG. 11, where ine magnet icons 306 appear to be pulling towards the fourth selection target 908, but clue to the faction selection criteria only four magnet icons 306 are actually included in the fourth faction 920,
  • FIG. 12 a flowchart of a novel method for performing faction analysis is shown. As described further below, the method of FIG. 12 may be applied in a first embodiment, described with respect to FIG. 13.
  • FIG. 13 shows the plurality of time step target areas 900, 1000, 1100 showing faction-associated spatial regions at three different time steps of the exemplary decision period, in accordance with the first embodiment of the method of faction determination of FIG. 12,
  • FIG. 14 shows a second embodiment of the method of generating faction data which can more accurately reflect faction represe tation using a novel method.
  • FIG. 12 a flowchart for a method of faction determination is shown in FIG. 12.
  • the target areas 900, 1000, 1100 of FIGS. 9-11 are shown including faction spatial regions in accordance with the first embodiment of the method of faction determination of FIG. 12. Shown are the pointer 210, the plurality of magnet icons 306, the answer choices 208, the plurality of selection targets 902, 904, 906, 908, 910, 912, a first Faction Associated Spatial Region
  • FASR FASR 1300
  • second FASR 1302 a third FASR 1304, a fourth FASR 1306, a fifth FASR 1308, a sixth FASR 1310, and a plurality of vertex angles 1312.
  • the method for determining faction analysis data is described wherein the CCS 102 is configured to define a plurality of spatial regions, each spatial region
  • FASRs Faction Associated Spatial Regions
  • each FASR is indexed with respect to its associated faction (answer choice) .
  • Each FASR is shaped as a convex angular slice with the vertex of the angular slice at the center of the graphical puck 210, with the angular slice orientation such that the open end opposite to the vertex is centered on the selection target associated with that faction.
  • the size of the vertex angle 1312 is approximately 30 degrees, i.e.
  • the exemplary target area shown in FIG. 13, ' which has six possible answer choices 208 (and corresponding selection targets 902, 904, 906, 908, 910, 912) and thus six factions, as a result there are six FASRs, each FASR associated with a faction that is pulling towards one of selection targets 902, 904, 906, 908, 910, 912 associated with the answer choices 208, denoted in FIG. 13 as "1", "2", “3", "4", "5", and w 6".
  • the FASR corresponding to answer choice 208 "1" is denoted the first FASR 1300, with the second FASR corresponding to answer choice 208 "2", etc.
  • first time step target area 900 shows the fifth FASR 1308, which is an angular slice of approximately 30 degrees with its vertex at the center of the graphical pointer 210 and orientated aimed such that the open angle is centered directly on answer choice "5", which is the answer choice associated with faction 5.
  • the other five FASRs 1300, 1302, 1304, 1306, 1310 are shown as well, pointing towards a swer choice 208 "1", answer choice 208 "2", answer choice 208 "3", answer choice 208 "4", answer choice 208 "5" and answer choice 208 "6" respectively.
  • the CCS 102 determines the location of the pointer 210 for the current time step.
  • the first time step target area 900 of FIG. 13 snows the first time step target area 900 at the time step occurring at 0 seconds, i.e. at the start of the decision period.
  • the pointer 210 is located at the pointer start point, equidistant from all selection targets 902, 904, 906, 908, 910, 912.
  • the FASR is determined for each answer choice .
  • the six FASRs 1300, 1302, 1304, 1306, 1308, 1310 are defined as the six angular regions (shown as shaded areas in FIG. 13), each of the angular regions having an origin at the center of the pointer 210 and aimed towards its
  • the vertex angle 1312 at each FASR vertex is the same, and is approximately 30 degrees in the embodiment shown for all factions at all time steps, while the orientation of each FASR is
  • the 102 determines, for each FASR, which magnet icons 306 are located within the FASR.
  • the definition of "within” may vary.
  • the magnet icon 306 is defined to be 'within the FASR when a centerline of the icon falls within the edges of the FASR.
  • Magnet icons 306 are used in the examples shown, but it will be understood that any suitable type of icon or other location indicator may be used.
  • the CCS 102 software counts 1 user in Faction 1, for there is one magnet icon 306
  • the CCS 102 software counts 1 user in Faction 3 (the third FASR 1304), and 1 user in Faction 5 (the fifth FASR 1308), for there is one magnet icon 306
  • the CCS 102 software counts 2 users in Faction 6 associated with the answer choice 208 "6". That's because two magnet icons 306 shown have their centerline within the limits of the sixth FASR 1310.
  • the CCS 102 software counts the number of users in each of the six factions at time step 0.
  • next time step 1206 the next time step is reached and the process returns to the pointer location step, where the pointer location is re ⁇ calculated.
  • the CCS 102 software is configured to repeat this method at each of a plurality of time steps. In most embodiments, time steps are every half-second, but in a preferred embodiment, the CCS 102 software uses time steps equal to a quarter second. Thus, four times per second, across the decision period, the CCS 102 software de ermines the number of users present in each of the plurality of factions based upon the User Intent Vectors for all users during that time-step, and the re-computed FASRs for that time step. A time-history of faction counts is stored by the CCS 102 software, indicating for each time step the number of users in each faction.
  • the CCS 102 software defines and represents the FASRs at each time- step during the decision period, we can now describe how these FASRs are used in the faction analysis. More specifically, for each time step during the decision period, the CCS 102 software is configured to count the number of users who are pulling on the puck 210 with a User Intent Vector that falls inside the boundaries of the angular range defined by the FASR indexed with each answer choice 208.
  • the puck 210 moves under collective control.
  • the FASRs 1300, 1302, 1304, 1306, 1308, 1310 are recomputed by the CCS 102 software, adjusting the origin of each FASR 1300, 1302, 1304, 1306, 1308, 1310 to the updated center location of the puck 210, and updating the orientation of each such that it continues to point at its respective selection target 902, 904, 906, 908, 910, 912,
  • second time step target area 1000 of FIG. 13 we see that after six seconds the puck 210 has moved to the new location. Consequently, the origin and orientation of each FASR 1300, 1302, 1304, 1306, 1308, 1310 has been recomputed by the CCS 102 software.
  • this time step corresponding to an elapsed time of 6 seconds into the decision process.
  • the magnet icons 306 and thus the User Intent Vectors
  • the FASRs 1300, 1302, 1304, 1306, 1308, 1310 have changed their positions and orientations, resulting from the new position of the pointer 210.
  • the CCS 102 software is configured to determine faction counts for this new configuration. In this example, the CCS 102 software would count 0 users in the sixth FASR 1310 corresponding to the faction
  • FIG. 13 depicts the new fifth FASR 1308 origin and orientation as defined by CCS 102 software for the given time-step.
  • the CCS 102 software re-computes the origin and
  • the fourth FASR 1306 spans the same angular area, but because the pointer 210 is close to the fou h selec ion target 908, the method is being less inclusive. This is a problem because human users adapt their strategy as the puck 210 nears an answer choice, widening the angle of pull. When a puck is far from an answer choice, human users tend to pull in an angle that is aimed at the answer choice within a narrow band, but as the puck approaches the answer choice, human users tend to 'widen the band they are pulling within even though their intent is still to get the puck to that answer choice. Because of this variability in human user angular alignment, an enhanced method is required to count factions which addresses the unique behavior of human users .
  • the first time step target area 900, the second time step target area 1000, and the third time step target area 1100 of FIGS . 9-11 are shown including FASRs 1300, 1302, 1304, 1306, 1308, 1310 during time steps of the decision process in
  • the time step target areas 900, 1000, 1100 of FIG. 14 represent an enhanced method of Faction Analysis.
  • the embodiment still uses the method of FIG. 12, with FASRs and faction membership determined by the CCS 102
  • each FxASR updated at each time step such that the FASR origin corresponds to the center of the puck 210 and the orientation of each FASR aims at a corresponding answer choice 208 at that time-step.
  • the counting process is also the same, such that at each time step, the CCS 102 software counts the number of users whose User Intent Vector (i.e. the centerline of the magnet icon) falls within each FASR, counting the users as contributors o a corresponding faction.
  • the vertex angle 1312 of each FASR is not fixed as in the first embodiment, but is varied by the CCS 102 software during the decision period based on the proximity of the pointer 210 to the corresponding answer choice 208. More
  • the FASR associated with the selection target (and therefore the associated answer choice 208) is increased in angular size (i.e. the vertex angle 1312 is increased) by the CCS 102 software as the pointer 210 approaches that particular selection target.
  • the FASR associated with one answer choice 208 is decreased in angular size (i.e. the vertex angle 1312 is decreased) by the CCS 102 software as the pointer 210 moves away from that particular answer choice 208.
  • the CCS 102 software is configured, at each time step, to compute the distance between the center of the puck 210 and each of the selection targets, and ad ust the vertex angle 1312 of the FASR associated with the selection target/answer choice 208 based on the distance to that answer choice 208.
  • One specific vertex calculation function is described below in FIGS . 15 and 16.
  • each FASR vertex angle 1312 is the same, as shown in the first time step target area 900.
  • the pointer 210 has been moved by the collective input to the location closer to the fourth selection target 908 corresponding to answer choice 208 "4".
  • the distance between the pointer 210 and the third selection targe 906, the fourth selection target 908, and the fifth selection target 910 has decreased, and as a result the vertex angles 1312 for the third FASR 1304, the fourth FASR 1306, and the fifth FAS . 1308 have widened.
  • the distance between the pointer 210 and the first selection target 902, the second selection target 904, and the sixth selection target 912 has increased, and as a result the vertex angles 1312 for the first FASR 1300, the second FASR 1302, and the sixth FASR 1310 have narrowed.
  • the pointer 210 has been moved by the collective input to the location even closer to the fourth selection target 908, nearly selecting the target 908.
  • the distance between the poin er 210 and the fourth selection target 908 has decreased further, and as a result the vertex angle 1312 for the fourth FASR 1306 has widened compared to the second time step target area 1000.
  • the distance between the pointer 210 and the first selection target 902, the second selection target 904, the third selection target 906, the fifth selection target 910, and the sixth selection target 912 has increased, and as a result the vertex angles 1312 for the first.
  • the FASR 1300, the second FASR 1302, the third FASR 1304, the fifth FASR 1308, and the sixth FASR 1310 have narrowed compared to the second time step target area 1000.
  • the vertex angles 1312 of the third FAS . 1304 and the fifth FASR 1308 are similar to the original vertex angles 1312 at the first time step.
  • the vertex angles 1312 of the first FASR 1300, the second FASR 1302, and the sixth FASR 1310 narrowed over both time steps, and are generally narrower than at the first time step.
  • the widening of the fourth FASR 1306 vertex angle 1312 as the pointer 210 approaches the fourth selection target 908 allows a more accurate accounting of the number of users exerting influence in the direction, of answer choice 208 "4". While as shown in FIG. 13 only four of the nine magnet icons 306 pulling on the pointer 210 are included in the fourth FASR 1306, using the second embodiment approximately 8 of the 9 magnet icons 306 are included in the fourth FAS . 1306, Referring next to FIGS . 15 and 16, spatial region (FASR) diagrams of a portion of the first time step target area 900 and the second time step target area 1000 are shown for the second embodiment of the method of FIG. 12 . Only the first FASR 1300, the fourth FASR 1306, and the fifth FASR are for clarity. Also shown are a first FASR starting distance 1500, a first FASR current
  • the enhanced second FASR distance 1502 a fourth FASR starting distance 1504, a fourth FASR current distance 1506, a. fifth FASR starting distance 1508, a fifth FASR current distance 1510, and a target cent.er 1512.
  • the enhanced second FASR starting distance 1504 a fourth FASR starting distance 1504 a fourth FASR current distance 1506, a. fifth FASR starting distance 1508, a fifth FASR current distance 1510, and a target cent.er 1512.
  • FASR calculation is as follows for each FASR correspondi g to one answer choice 208. The following values are defined:
  • Direction the vector from the center of the pointer to the center of the selection target.
  • Starting Distance (S) distance from, center of the pointer to center of selection target at start time.
  • the vertex angle (1312 angular size) of each FASR is then calculated as:
  • Vertex angle 20° + 40°*S/ (5*C) ⁇ 60°
  • the vertex angle 1312 is defined as having a maximum value not to exceed 60 degrees .
  • each starting distance S 1500, 1504, 1508 equals the
  • the corresponding current distance C 1502, 1506, 1510, and the vertex angles 1312 are all the same (28 degrees using the present formula) .
  • the pointer 210 has moved away from the first selection target 902, closer to the fifth selection target 910, and even closer to the fourth selection, target 908.
  • the current distances 1502, 1506, 1510 are as shown on FIG. 16, with the first FASR current distance 1502 longer than the first FASR starting distance 1500, the fourth FASR current distance 1506 shorter than the fourth FASR starting distance 1504, and the fifth FASR current distance 1510 also shorter than the fifth FASR starting distance 1508.
  • These current distances 1502, 1506, 1510, * when input into the formula result in the smaller vertex angle 1312 for the first FASR. 1300 and the larger vertex angles 1312 for the fourth FASR 1306 and the fifth FASR 1308.
  • the angular region (vertex angle 1312) of the FASR that's associated with that answer choice will decrease linearly as the current distance grows.
  • the angular size will drop from the staring angle of 28 degrees, linearly approaching 20 degrees as distance rises. How close it gets to 20 degrees depends on the size of the FASR involved.
  • This unique processing by the CCS 102 software accounts for the fact that human users express their intent with more angular precision as the target gets further away, thus falling within a narrower band when, expressing an intent for a particular target . Without this processing, the CCS 102 software might include a user in a faction that does not correspond with his or her actual intent.
  • the CCS 102 software is configured such that, as the puck 210 moves towards a particular selection target, the angular region (vertex angle 1312) of the FASR associated with that selection target will grow linearly as the current distance shrinks.
  • the angular size will grow from the starting angle of 28 degrees, getting larger and larger as the distance shrinks, until the value is capped at 60 degrees by the CCS 102 software (as mentioned above) , Note, in some embodiments, the cap could be higher than 60 degrees.
  • the intent of the cap is to avoid overlap of FASR regions which would be indeterminate.
  • a linear relation is used herein, other relations between distance and angular size may be employed. Also, if a linear relation is used, other slopes may be used for that relation.
  • This unique processing by the CCS 102 software accounts for the fact that human users express their intent with less angular precision as the target gets closer, thus falling within a wider band when expressing an intent, for a particular target. Without this processing, the CCS 102 software might fail to include a user in a faction that
  • the relationship between the distance and angular size is non- linear, the profile crafted to more accurately match human behavioral tendencies when pulling towards one target. More
  • the range of possible angles used by users when pulling towards the target expands particularly rapidly when the pointer gets very close to the target.
  • This can inventively be modeled as a power function where the angle size increases proportional to the distance raised to the -1.3 power,
  • the CCS 102 software determines the number of users pulling towards each of the plurality of selection targets (corresponding to answer choices 208) at each of the plurality of time-steps across the
  • the CCS 102 software is
  • the CCS 102 software can be configured to sum the magnitudes of the force magnitudes associated with the user intent vectors that fall into a particular faction at a particular moment in time. This is referred to herein as a "force summation" based Faction Analysis.
  • the total force across all factions is not necessarily the same as the total force on the puck 210, for some users pull between factions, their intent not aimed at any particular answer at a moment in time. These are users who are pulling in the areas between FASRs .
  • the total force applied between factions is also summed and tracked across time steps.
  • the count of users between factions and/or the summation of user force between factions is used to indicate an intention of users to "defend against" an answer choice 208 the puck 210 is heading towards, such a defense not being associated with a particular alternative answer. It is often the case that users vary their strategy during a collective intelligence decision, varying between
  • the question display 302 Shown are the question display 302, the plurality of answer choices 208, a plurality of selection targets 1702 (also wherein the plurality of selections targets 1702 comprises the individual selection targets 902, 904, 906, 908, 910, 912), the pointer 210, and the plurality of magnet icons 306.
  • the CCS 102 controls the motion of the puck 210 based on the
  • the puck 210 is collectively controlled across the decision period, moved from a starting position to the final target selection by the plurality of users 'working together in real-time as a unified dynamic system.
  • the CCS 102 (or related subset of software) is configured to determine which of the six answer choices 208 each user is pulling towards, at each time step during the decision process, or if any users are not pulling towards any faction. In this case the six factions correspond with the six candidate names, which are the set of possible answers presented to the collective
  • the faction analysis data thus describes in an. efficient and understandable way, how user factions form and change and sometimes dissolve across the
  • FIG. 18 an example time-history plot is shown that depicts the time-history of faction data across the decision period associated with the question of FIG. 17. Shown are a Ted Cruz faction data line 1800, a Scott Walker faction data line 1802, a Marco Rubio faction data line 1804, a Jeb Bush data line 1806, a Mike Huckabee data line 1808, and a Chris Christie data line 1810.
  • FIG. 17 shows a radial plot that starts at the top (i.e. at twelve o'clock) at 0 seconds, then proceeds clockwise around the circle, conveying a full decision period that lasts approximately 30 seconds.
  • each data line 1800, 1802, 1804, 1806, 1808, 1810 plotted on the chart indicates how many users are included in each faction at the given time.
  • the plot shows, using a different linetype for each faction, the number of users present in each of the factions over time.
  • the height of each data line 1800, 1802, 1804, 1806, 1808, 1810 shows the number of users present in the faction associated with that line 1800, 1802, 1804, 1806, 1808, 1810.
  • the key on the chart shows which linetype is associated with each candidate faction .
  • the height of the Jeb Bush data line 1806 peaks at 5 users, all pulling together towards Jeb Bush.
  • the users were able to get the puck 210 onto the Jeb Bush associated answer choice 208.
  • the collective intelligence was able to find common ground, converging on Jeb Bush as the answer.
  • the inventive process of faction analysis described herein provides new and important insights into the decision process of a plurality of users making decisions as a real-time co 11 ective inte11igence .
  • the present invention also includes a method for quantifying the behavior of each of the plurality of users based on the faction analysis data. More specifically, the present invention is configured such that the time-history of faction data records which user was present in which faction at each time step. This is achieved by the CCS 102 storing a unique user identifier for each of the plurality of users and associating that identifier with the data stored in the faction time history. In this way, the CCS 102 stores an indication of which user was present in which faction at each ime step during the decision period and if so, how many times they changed. This data can be processed on a per-user basis, determining if that user changed factions during the decision period. If a user did not change factions at all during the decision period, that user is
  • the user changed factions a small number of times during the decision process (i.e. the number of times the user changed factions is between an upper limit and a lower limit), for example between two and four times, that user is classified as "flexible" by the CCS 102 software process. And if the user changed factions a large number of times (i.e. the number of times the user changed factions is larger than an upper limit) , for example, five or more times during the decision process, that user is classified as "fickle" by the CCS 102 software process . In this way participants in a real ⁇ time collective intelligence can be assessed based on the number of times they changed factions during a real-time collective decision process. In some embodiments the user is classified based on a number of different
  • users are awarded points or credits for being classified as flexible, but awarded less points or credits (or no points or credits) if classified as entrenched or fickle.
  • users who are classified by the CCS 102 software as entrenched and/or fickle lose points or credits . This encourages effective decision making as a unified intelligence.
  • swarm intelligence systems enable a group of users to express their opinion in real time, varying both their choice and their expressed level of conviction smoothly during the decision period.
  • The enables the closed loop system that is a real-time physical negotiation between users .
  • each user expresses their opinion (both the option and the level of
  • the direction and magnitude of the user input is the user intent vector.
  • FIG. 17 shows the display of an exemplary
  • each magnet icon 306 is controlled by an individual user, said magnet icons 306 applying simulated forces upon the
  • the closed loop system of users thereby controls the pointer 210 to answer a question as a real-time system, enabling the formation of an emergent group intelligence.
  • the present invention improves upon these prior art systems and methods by creating what is referred to as a "Smart Magnet” user input that varies its impact upon the real-time physical system based upon predictive
  • the present invention provides a non-linear user intent vector with a simulated physical "pull" on the graphical pointer 210 that is not only a function of the relative direction and magnitude of the user intent vector (as represented graphically by the orientation of the magnet icon 306 and the proximity of the magnet icon 306 to the pointer 210), as in prior systems, but also a function of the time-history of the user intent vector during a current group decision process. More specifically, the present invention reduces the pull strength (magnitude of the user intent vector) for a particular user for a period of time following that user changing their pull direction (direction of the use - intent vector) from supporting the motion towards one selection target 1702, to supporting the motion of the pointer 210 towards a different selection target 1702. This reduction in under input vector magnitude (pull strength) is referred to herein as attenuation and is controlled by software algorithms of the CCS 102, said algorithms referred to herein as attenuation algorithms.
  • the algorithm is configured run locally on the portable computing device 104 of each user as part of the CIA application that communicates with the CCS 102.
  • the CIA communicates attenuated user intent vectors to the CCS 102 for the user of that portable computing device 104.
  • This methodology has the advantage of distributing the processing load, such that the plurality of portable computing devices 104 each share the processing burden of the attenuation algorithm for their given users .
  • a first method for adjusting the pull strength of a user based on changing of pull direction is shown. Shown are a determine current selection target step 1900, a changed selection target decision point 1902, an increment value step 1904, and an ad ust magnitude step 1906.
  • the CCS 102 or CIA determines, using the
  • the real-time faction analysis algorithms are performed by the CCS 102, for example when the attenuation algorithms are run on the CCS 102. In other embodiments the faction analysis algorithms are performed by the CIA, for example 'when the attenuation algorithms are distributed on the portable computing devices 104.
  • the CCS 102 (or CIA) compares the currently
  • the method returns to the determine current selection target step 1900, and the CCS 102 (or CIA) checks the se1ection target 1702 again .
  • the CCS 102 keeps track of how many times the user has changed selection targets 1702 during the session by incrementing a value associated with the number of selection target changes. For example, if the change is the first change during the session, the CCS 102 (or CIA ⁇ increments the value from. 0 to 1.
  • a time threshold is implemented such that the user intent vector is not determined to have changed factions (i.e. is not determined to be pulling towards a different selection target 1702) until it's within that faction for more than a threshold amount of time. In some embodiments, the time threshold is between 250 and 750 milliseconds.
  • This time frame is chosen so as to not count factions that a user is simply passing through, but factions a user is settling within. For example, if a user is pulling for first selection target 902 (i.e. is determined to be included in the FASR associated with the first selection target 902) and switches to third selection target 906, they might move their vector through the FASR associated with second selection target 904. If this "pass through” occurs in under the threshold amount of time, for example under 500 milliseconds, it is not counted a change and thus does not increment the count. Conversely, when the user settles in the FASR associated with the third selection target 906 for more than the 500 millisecond threshold, it is counted as a change and does increment the count.
  • the CCS 102 (or CIA) adjusts the magnitude of the user intent vector based on the number and/or timing of the selection target changes over a period of time.
  • the attenuation algorithm reduces the user intent vector magnitude for each user for a period of time based on the number of selection target changes over a period of time executed by that user.
  • the magnitude may be configured to drop by an attenuation amount, then ramp smoothly back to full magnitude over a restoration period.
  • restoration period may be dependent upon the number of selection target changes during a particular time period. For example, the more selection target changes during the time pe iod, the greater the attenuation amount and/or the greater the restoration period.
  • the algorithm reduces the magnitude of the user intent vector based on how much time has passed since the last target change.
  • the magnitude may be configured to drop by an attenuation amount, then ramp smoothly back to full magnitude over a restoration period.
  • the size of the attention amount and/or the duration of the restoration period may be dependent upon the time that has passed since the last selection target change . For example, the shorter the time since the last selection target change, the greater the attenuation amount and/or the greater the restoration period.
  • the magnitude of the reduction in user intent vector magnitude goes up with the number of changes in direction from one selection target 1702 to a
  • selection target 1702 during a particular answer period. For example, if one user provides user input pulling towards selection target 1, then later the user swi ches t.o pu11ing towards se1ection target 2 , t.he attenuation algorithm will then temporarily reduce the user intent vector magnitude by Attenuation Percentage A. Then, if the same user changes their pull direction to a different selection target (or back to selection target 1) , that user's user intent vector will then temporarily reduce from the user' s original full intent vector magnitude by Attenuation Percentage B, which is greater than Attenuation Percentage A.
  • selection target 1702 he or she is pulling for during a real-time group decision will be provided by the methods herein with a user intent vector magnitude that is greater compared to a user who changes input direction during the session.
  • a user who changes the selection target 1702 he is pulling for more frequently during a real-time decision will have greater attenuation of user input vector magnitude (magnet strength) than a user who changes the selection target 1702 he is pulling for less frequently. For example, a user who changes the selection target 1702 he is pulling for 4 times during a decision process will have a greater reduction
  • the reason for this attenuation based on change of selection target 1702 is that the change is used as predictive indicator of lower conviction level by the user who makes the change.
  • a user who changes 5 times is predicted to nave lower conviction in the selection target 1702 he is pulling for than a user who changes 2 times .
  • the distance of the magnet icon 306 from the pointer 210 also controls the conviction level, as described in the related patent applications, such that the present invention provides a predictive attenuation on the user-controlled conviction, said predictive attenuation being based on changes in the se1ection target 1702 being pu11ed for .
  • the reduction in user input vector magnitude, as adjusted by the Attenuation Algorithm is a momentary drop in magnitude followed by a gradual ramp up in magnitude until the user intent vector- magnitude returns to the unaffected condition (e.g. "full strength") .
  • the ramp up is linear. In other preferred embodiments, the ramp up is non-linear. In one preferred embodiment, the attenuation is approximately 25% of the normal full strength of the user intent vector magnitude.
  • the attenuation algorithm then gradually returns the user input vector to full strength (i.e. solely
  • the restoration period is 2 to 4 seconds.
  • 2 to 4 seconds is a rapid restoration, thus causing a subtle reduction that is not physically noticeable to the user controlling the magnet icon 306, but does change the relative impact of that user' s input with respect to the plurality of other users.
  • the simulated physical system will converge on an answer based at least in part upon on predictive conviction variations among users.
  • a further inventive method described herein includes algorithmic consideration for the overall motion of the pointer 210, assessing whether a user, upon changing the selection target 1702 they are pulling for, is then pulling WITH the direction of motion of the pointer 210, or AGAINST the direction of motion of the pointer 210. For example, if the pointer 210 is moving towards
  • selection target A at a moment in time i.e. the group i tent vector direction is towards selection target A
  • one user switches from pulling towards selection target B (i.e. the user input vector direction towards selection target B) to pulling towards selection target A (i.e. the user input vector direction towards selection target A)
  • that user would be algorithmically determined to have switched to pulling with, or supporting, the direction of the group intent vector, as represented by the current direction of the pointer 210.
  • the algorithmic methods determine if the pointer 210 is not currently moving towards any selection target 1702 at a moment in time. This is determined with a threshold velocity value of the pointer 210 motion, below which the pointer 210 is considered to be inconclusive in terms of which option the pointer 210 is moving towards. This creates three possible conditions that
  • Condition 1 the user switches to a direction substantially aligned with the current direction that the pointer 210 is currently moving in (i.e. the user intent vector direction is substantially aligned with the group intent vector direction) . This condition is referred to as "with the flow”, as the user's pull is now going with the direction of the pointer 210.
  • Condition 2 the user switches to a direction different from the direction in which the pointer 210 is moving. This condition is referred to as "against the flow" as the user' s pull is not
  • Condition 3 is referred to as "neutral flow" when there is not a clear directional motion of the pointer 210 towards a particular selection target 1702 at that moment in time, so that user's switch has no clear indication as to being with the flow or against the flow.
  • the inventive method disclosed herein uses that Flow Condition to modulate the attenuation profile of that user's user intent vector at that moment in time (i.e. at changing of selection target 1702) .
  • the inventive method disclosed herein modulates the degree of attenuation based on whether the user' s change in selection target 1702 was "with the flow", “against the flow” or “neutral flow”. This allows the attenuation to predictively take into consideration the impact of "Social Influence Bias", as a user who switches to pulling with the flow is more likely to have been influenced by other users, than a user who pulls against the flow. Said another way, a user who makes a change that goes with the majority of other users is more likely to have been influenced by Social
  • the degree of attenuation is modulated based on the flow magnitude. For example, a user 'who goes with the flow, when there is a high flow magnitude will have greater attenuation than a user who goes with the flow when there is low flow magnitude .
  • FIG. 20 a second method for adjusting the pull strength of a user based on condition is shown. Shown are a determine current selection target step 1900, a compare intent vectors step 2000, a
  • the CCS 102 determines, using the direction of the user intent vector, which selection target 1702 the user intent vector is directed towards, i.e. which selection target 1702 the user is "pulling towards” .
  • point 1900 the CCS 102 (or CIA) determines, using the direction of the user intent vector, which selection target 1702 the user intent vector is directed towards, i.e. which selection target 1702 the user is "pulling towards” .
  • the CCS 102 (or CIA) compares the currently
  • the me hod returns to the determine current
  • the method proceeds to the compare intent vectors step 2000.
  • the CCS 102 (or CIA) compares the direction of the user intent vector to the direction, of the velocity of the pointer 210.
  • determine flow condition 2002 the flow
  • condition is determined, i.e., for the new selection target 1702, whether the current user intent vector is "with the flow”, “against the flow”, or "neutral".
  • direction of the user intent vector is compared to the direction of the group intent vector (which the pointer velocity is based on)
  • the CCS 102 (or CIA) adjusts the magnitude of the user intent vector based on the determined flow condition.
  • the user intent vector is adjusted for a period of time. It will be understood that the methods of FIG . 19 and FIG. 20 may be used concurrently, i.e. the magnitude and/or period of time may be adjusted for number of changes and the flow condition at the same time. Referring again to FIG. 20, the inventive method herein modulates the attenuation profile of the
  • the attenuation in user intent vector magnitude is greater and/or the time to return to full magnitude is longer if that user is determined by the algorithm to have been influenced by social influence bias. Conversely, the attenuation in magnitude (strength) of a user's user intent vector magnitude is lower and/or the time to return to full magnitude is shorter if that user is predicted to have not been influenced by social influence bias.
  • control algorithms disclosed herein are formulated such that if a user changes the selection target 1702 they are pulling towards, and also does so "with the flow" of the pointer 210 (i.e. changes to the direction towards the selection target 1702 the group-controlied pointer 210 is currently moving
  • control algorithms disclosed herein are formulated such that if a user changes the selection target 1702 they are pulling towards, and also does so "against the flow" of the pointer 210 (i.e. changes to a direction different from the direction the pointer 210 is moving in) , the
  • the attenuation percentage is lower and/or the ramp-up period is shorter. In this way, a user who is more likely to be influenced by social influence bias has a reduced impact on the collaborati ely controlled motion of the pointer 210 as compared to a user who is less likely to be influenced by social influence bias. In addition, if after the change in direction the pointer 210 is
  • the contro1 a1 gori thms provide an a11enuat ion percentage and/or a ramp-up period that falls between the two extremes indicated by the "with the flow” and "against the flow” conditions.
  • a particular embodiment is configured such that the baseline attenuation percentage is 25% and the baseline ramp-up period is 3 seconds. In such an embodiment, if a user switches his pull from towards one selection target 1702 to a different selection target
  • the simulated pull of the user's magnet icon 306 (as a graphical representation of the user's user intent vector) on the collaboratively controlled pointer 210 is reduced by 25% of full strength, and ramps up linearly back to full strength over 3 seconds.
  • the simulated pull of the user's magnet icon 306 on the collaboratively controlled pointer 210 is reduced by 25% of full strength, and ramps up linearly back to full strength over 3 seconds.
  • the user intent vector is directed towards, during an interactive closed- loop collaborative decision process, the present
  • the piresent invention accounts predictively for the impact of social influence bias on user behavior. This creates a higher likelihood that the final result is reflective of the collective group wisdom.
  • the attenuation percentage may be selectively increased based also on the number of changes in selection target direction during a decision period.
  • the restoring period can be selectively increased based also on the number of changes in
  • the algorithmic methods by which to determine which of a plurality of options a user is pulling for at a moment is in time are those used by the technique called "Faction Analysis'" as disclosed in FIGS. 9-18.
  • FASR Fraction Associated Spatial Region
  • the CCS 102 determines the size and location of the FASR as sociated with each se1 ection target 1702 , and then determines which magnet icons 306 are associated 'with each FASR. If the user's magnet icon 306 (with the location and movement determined by the user intent vector) falls within the FASR associated with one
  • the faction analysis methods compare the selection target 1702 being pulled for by each magnet icon 306 over time, tracking if the user switches from pulling for one selection target 1702 (e.g. the magnet icon 306 is in the FASR associated with a first selection target) to pulling for a different selection target 1702 (e.g. the magnet icon 306 moves over time to a FASR associated with a second selection, target) .
  • the methods do this for all magnet icons
  • the displayed size of the magnet icon 306 on the screen of a user is modulated based on the attenuation level imposed by the algorithms herein.
  • an attenuated magnet icon 306 is displayed smaller than the comparative full strength magnet icon 306. This gives a user feedback as to the attenuation of their magnet icon 306.
  • the size is not attenuated during real-time display of the magnet icon 306 during the decision process, but is displayed in later replays of the magnet swarm. Replays for sessions are disclosed at least in the co-pending U.S. Patent.
  • the pointer 210 begins at a high mass value and gradually decreases over time. This makes it the most difficult for the users to influence the motion of the pointer 210 at the start of collaborative control, but gets easier to influence the motion of the pointer 210 later in the collaborative control.
  • a benefit of this inventive method is that it reduces the impact of users during the early period in which the users are making up their mind, and increases the impact of the users during a later period when users have more conviction in their opinions. Said another way, it ensures users have time to think before they act, because acting is harder at the start and gets easier over time as a result of this non-linear physical model.
  • This method which is referred to herein as variable physical dynamics, has been found to increase the performance accuracy of real-time human systems working as a unified intelligence.
  • modules may be implemented as a hardware circuit comprising custom VLSI circuits or gate arrays, off-the-shelf semiconductors such as logic chips, transistors, or other discrete components.
  • a module may also be implemented in
  • programmable hardware devices such as field programmable gate arrays, programmable array logic, programmable logic devices or the like.
  • Modules may also be implemented in software for execution by various types of processors.
  • An identified module of executable code may, for instance, comprise one or more physical or logical blocks of computer
  • a module of executable code could be a single instruction, or many instructions, and may even be distributed over several different code segments, among different programs, and across several memory devices. Similarly, operational data may be identified and
  • modules may be embodied in any suitable form and organized within any suitable type of data structure.
  • the operational data may be collected as a single data set, or may be distributed over

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Human Resources & Organizations (AREA)
  • Data Mining & Analysis (AREA)
  • Strategic Management (AREA)
  • Quality & Reliability (AREA)
  • Biomedical Technology (AREA)
  • Economics (AREA)
  • Marketing (AREA)
  • Operations Research (AREA)
  • Educational Administration (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Educational Technology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • User Interface Of Digital Computer (AREA)

Abstract

L'invention concerne des systèmes et des procédés de calcul collaboratif et d'intelligence collective en temps réel. Une application collaborative s'exécute sur un serveur collaboratif connecté à une pluralité de dispositifs informatiques. Des sessions collaboratives sont exécutées, dans lesquelles un groupe d'utilisateurs indépendants, reliés en réseau sur l'internet, répondent de manière collaborative à des questions en temps réel, ceci permettant d'exploiter leur intelligence collective. L'invention porte également sur des procédés qui permettent, pendant une session collaborative, de modifier en temps réel la valeur quantitative de l'influence que l'utilisateur a dans la détermination de la réponse collaborative.
PCT/US2017/040480 2014-03-26 2017-06-30 Procédés et systèmes pour modifier l'influence de l'utilisateur pendant une session collaborative d'un système d'intelligence collective en temps réel Ceased WO2018006065A1 (fr)

Priority Applications (10)

Application Number Priority Date Filing Date Title
US15/815,579 US10439836B2 (en) 2014-03-26 2017-11-16 Systems and methods for hybrid swarm intelligence
PCT/US2017/062095 WO2018094105A2 (fr) 2016-11-17 2017-11-16 Systèmes et procédés d'intelligence collective hybride
US15/898,468 US10712929B2 (en) 2014-03-26 2018-02-17 Adaptive confidence calibration for real-time swarm intelligence systems
US15/904,239 US10416666B2 (en) 2014-03-26 2018-02-23 Methods and systems for collaborative control of a remote vehicle
US15/922,453 US20180204184A1 (en) 2014-03-26 2018-03-15 Parallelized sub-factor aggregation in real-time swarm-based collective intelligence systems
US16/059,698 US11151460B2 (en) 2014-03-26 2018-08-09 Adaptive population optimization for amplifying the intelligence of crowds and swarms
US16/154,613 US11269502B2 (en) 2014-03-26 2018-10-08 Interactive behavioral polling and machine learning for amplification of group intelligence
US16/230,759 US10817158B2 (en) 2014-03-26 2018-12-21 Method and system for a parallel distributed hyper-swarm for amplifying human intelligence
US17/237,972 US11636351B2 (en) 2014-03-26 2021-04-22 Amplifying group intelligence by adaptive population optimization
US18/114,954 US12099936B2 (en) 2014-03-26 2023-02-27 Systems and methods for curating an optimized population of networked forecasting participants from a baseline population

Applications Claiming Priority (8)

Application Number Priority Date Filing Date Title
US15/199,990 2016-07-01
USPCT/US2016/040600 2016-07-01
US15/199,990 US20160314527A1 (en) 2014-03-26 2016-07-01 Methods and systems for enabling a credit economy in a real-time collaborative intelligence
PCT/US2016/040600 WO2017004475A1 (fr) 2015-07-01 2016-07-01 Procédés et systèmes pour permettre une économie de crédit dans une intelligence collaborative en temps réel
US201662358026P 2016-07-03 2016-07-03
US62/358,026 2016-07-03
US15/241,340 US10222961B2 (en) 2014-03-26 2016-08-19 Methods for analyzing decisions made by real-time collective intelligence systems
US15/241,340 2016-08-19

Related Parent Applications (5)

Application Number Title Priority Date Filing Date
US15/241,340 Continuation-In-Part US10222961B2 (en) 2014-03-26 2016-08-19 Methods for analyzing decisions made by real-time collective intelligence systems
US15/640,145 Continuation-In-Part US10353551B2 (en) 2014-03-26 2017-06-30 Methods and systems for modifying user influence during a collaborative session of real-time collective intelligence system
PCT/US2017/062095 Continuation-In-Part WO2018094105A2 (fr) 2014-03-26 2017-11-16 Systèmes et procédés d'intelligence collective hybride
US16/059,698 Continuation US11151460B2 (en) 2014-03-26 2018-08-09 Adaptive population optimization for amplifying the intelligence of crowds and swarms
US17/237,972 Continuation US11636351B2 (en) 2014-03-26 2021-04-22 Amplifying group intelligence by adaptive population optimization

Related Child Applications (8)

Application Number Title Priority Date Filing Date
US14/668,970 Continuation-In-Part US9959028B2 (en) 2014-03-26 2015-03-25 Methods and systems for real-time closed-loop collaborative intelligence
PCT/US2016/040600 Continuation-In-Part WO2017004475A1 (fr) 2014-03-26 2016-07-01 Procédés et systèmes pour permettre une économie de crédit dans une intelligence collaborative en temps réel
US15/640,145 Continuation-In-Part US10353551B2 (en) 2014-03-26 2017-06-30 Methods and systems for modifying user influence during a collaborative session of real-time collective intelligence system
US15/815,579 Continuation-In-Part US10439836B2 (en) 2014-03-26 2017-11-16 Systems and methods for hybrid swarm intelligence
PCT/US2017/062095 Continuation-In-Part WO2018094105A2 (fr) 2014-03-26 2017-11-16 Systèmes et procédés d'intelligence collective hybride
US16/059,698 Continuation-In-Part US11151460B2 (en) 2014-03-26 2018-08-09 Adaptive population optimization for amplifying the intelligence of crowds and swarms
US16/154,613 Continuation-In-Part US11269502B2 (en) 2014-03-26 2018-10-08 Interactive behavioral polling and machine learning for amplification of group intelligence
US16/230,759 Continuation-In-Part US10817158B2 (en) 2014-03-26 2018-12-21 Method and system for a parallel distributed hyper-swarm for amplifying human intelligence

Publications (1)

Publication Number Publication Date
WO2018006065A1 true WO2018006065A1 (fr) 2018-01-04

Family

ID=60787684

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US2017/040480 Ceased WO2018006065A1 (fr) 2014-03-26 2017-06-30 Procédés et systèmes pour modifier l'influence de l'utilisateur pendant une session collaborative d'un système d'intelligence collective en temps réel

Country Status (1)

Country Link
WO (1) WO2018006065A1 (fr)

Cited By (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10606463B2 (en) 2014-03-26 2020-03-31 Unanimous A. I., Inc. Intuitive interfaces for real-time collaborative intelligence
US10606464B2 (en) 2014-03-26 2020-03-31 Unanimous A.I., Inc. Methods and systems for gaze enabled collaborative intelligence
US10609124B2 (en) 2014-03-26 2020-03-31 Unanimous A.I., Inc. Dynamic systems for optimization of real-time collaborative intelligence
US10656807B2 (en) 2014-03-26 2020-05-19 Unanimous A. I., Inc. Systems and methods for collaborative synchronous image selection
US11151460B2 (en) 2014-03-26 2021-10-19 Unanimous A. I., Inc. Adaptive population optimization for amplifying the intelligence of crowds and swarms
US11269502B2 (en) 2014-03-26 2022-03-08 Unanimous A. I., Inc. Interactive behavioral polling and machine learning for amplification of group intelligence
US11360656B2 (en) 2014-03-26 2022-06-14 Unanimous A. I., Inc. Method and system for amplifying collective intelligence using a networked hyper-swarm
US11360655B2 (en) 2014-03-26 2022-06-14 Unanimous A. I., Inc. System and method of non-linear probabilistic forecasting to foster amplified collective intelligence of networked human groups
US11941239B2 (en) 2014-03-26 2024-03-26 Unanimous A.I., Inc. System and method for enhanced collaborative forecasting
US11949638B1 (en) 2023-03-04 2024-04-02 Unanimous A. I., Inc. Methods and systems for hyperchat conversations among large networked populations with collective intelligence amplification
US12001667B2 (en) 2014-03-26 2024-06-04 Unanimous A. I., Inc. Real-time collaborative slider-swarm with deadbands for amplified collective intelligence
US12079459B2 (en) 2014-03-26 2024-09-03 Unanimous A. I., Inc. Hyper-swarm method and system for collaborative forecasting
US12099936B2 (en) 2014-03-26 2024-09-24 Unanimous A. I., Inc. Systems and methods for curating an optimized population of networked forecasting participants from a baseline population
US12190294B2 (en) 2023-03-04 2025-01-07 Unanimous A. I., Inc. Methods and systems for hyperchat and hypervideo conversations across networked human populations with collective intelligence amplification

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090063379A1 (en) * 2007-03-06 2009-03-05 Patrick Laughlin Kelly Automated decision-making based on collaborative user input
US20120011006A1 (en) * 2010-07-09 2012-01-12 Richard Schultz System And Method For Real-Time Analysis Of Opinion Data
WO2014023432A1 (fr) * 2012-08-09 2014-02-13 Livestudies Srl Appareil et procédé destinés à un environnement collaboratif utilisant des surfaces tactiles réparties
US20140379439A1 (en) * 2005-10-28 2014-12-25 International Business Machines Corporation Aggregation of subsets of opinions from group collaborations
US20150331601A1 (en) * 2014-03-26 2015-11-19 Unanimous A.I. LLC Intuitive interfaces for real-time collaborative intelligence

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20140379439A1 (en) * 2005-10-28 2014-12-25 International Business Machines Corporation Aggregation of subsets of opinions from group collaborations
US20090063379A1 (en) * 2007-03-06 2009-03-05 Patrick Laughlin Kelly Automated decision-making based on collaborative user input
US20120011006A1 (en) * 2010-07-09 2012-01-12 Richard Schultz System And Method For Real-Time Analysis Of Opinion Data
WO2014023432A1 (fr) * 2012-08-09 2014-02-13 Livestudies Srl Appareil et procédé destinés à un environnement collaboratif utilisant des surfaces tactiles réparties
US20150331601A1 (en) * 2014-03-26 2015-11-19 Unanimous A.I. LLC Intuitive interfaces for real-time collaborative intelligence

Cited By (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11636351B2 (en) 2014-03-26 2023-04-25 Unanimous A. I., Inc. Amplifying group intelligence by adaptive population optimization
US11769164B2 (en) 2014-03-26 2023-09-26 Unanimous A. I., Inc. Interactive behavioral polling for amplified group intelligence
US10609124B2 (en) 2014-03-26 2020-03-31 Unanimous A.I., Inc. Dynamic systems for optimization of real-time collaborative intelligence
US10656807B2 (en) 2014-03-26 2020-05-19 Unanimous A. I., Inc. Systems and methods for collaborative synchronous image selection
US11151460B2 (en) 2014-03-26 2021-10-19 Unanimous A. I., Inc. Adaptive population optimization for amplifying the intelligence of crowds and swarms
US11269502B2 (en) 2014-03-26 2022-03-08 Unanimous A. I., Inc. Interactive behavioral polling and machine learning for amplification of group intelligence
US10606464B2 (en) 2014-03-26 2020-03-31 Unanimous A.I., Inc. Methods and systems for gaze enabled collaborative intelligence
US11360656B2 (en) 2014-03-26 2022-06-14 Unanimous A. I., Inc. Method and system for amplifying collective intelligence using a networked hyper-swarm
US10606463B2 (en) 2014-03-26 2020-03-31 Unanimous A. I., Inc. Intuitive interfaces for real-time collaborative intelligence
US11360655B2 (en) 2014-03-26 2022-06-14 Unanimous A. I., Inc. System and method of non-linear probabilistic forecasting to foster amplified collective intelligence of networked human groups
US11941239B2 (en) 2014-03-26 2024-03-26 Unanimous A.I., Inc. System and method for enhanced collaborative forecasting
US12099936B2 (en) 2014-03-26 2024-09-24 Unanimous A. I., Inc. Systems and methods for curating an optimized population of networked forecasting participants from a baseline population
US12001667B2 (en) 2014-03-26 2024-06-04 Unanimous A. I., Inc. Real-time collaborative slider-swarm with deadbands for amplified collective intelligence
US12079459B2 (en) 2014-03-26 2024-09-03 Unanimous A. I., Inc. Hyper-swarm method and system for collaborative forecasting
US11949638B1 (en) 2023-03-04 2024-04-02 Unanimous A. I., Inc. Methods and systems for hyperchat conversations among large networked populations with collective intelligence amplification
US12190294B2 (en) 2023-03-04 2025-01-07 Unanimous A. I., Inc. Methods and systems for hyperchat and hypervideo conversations across networked human populations with collective intelligence amplification

Similar Documents

Publication Publication Date Title
US10353551B2 (en) Methods and systems for modifying user influence during a collaborative session of real-time collective intelligence system
US10222961B2 (en) Methods for analyzing decisions made by real-time collective intelligence systems
WO2018006065A1 (fr) Procédés et systèmes pour modifier l'influence de l'utilisateur pendant une session collaborative d'un système d'intelligence collective en temps réel
US10122775B2 (en) Systems and methods for assessment and optimization of real-time collaborative intelligence systems
US10609124B2 (en) Dynamic systems for optimization of real-time collaborative intelligence
US10712929B2 (en) Adaptive confidence calibration for real-time swarm intelligence systems
US10656807B2 (en) Systems and methods for collaborative synchronous image selection
US10277645B2 (en) Suggestion and background modes for real-time collaborative intelligence systems
US10551999B2 (en) Multi-phase multi-group selection methods for real-time collaborative intelligence systems
US11769164B2 (en) Interactive behavioral polling for amplified group intelligence
US11360656B2 (en) Method and system for amplifying collective intelligence using a networked hyper-swarm
US20160154570A1 (en) Iterative suggestion modes for real-time collaborative intelligence systems
US10606463B2 (en) Intuitive interfaces for real-time collaborative intelligence
US11360655B2 (en) System and method of non-linear probabilistic forecasting to foster amplified collective intelligence of networked human groups
WO2018094105A2 (fr) Systèmes et procédés d'intelligence collective hybride
EP3210386A1 (fr) Systèmes et procédés pour l'analyse des performances et la modération d'une intelligence collaborative en temps réel multi-niveaux
US12079459B2 (en) Hyper-swarm method and system for collaborative forecasting
US12001667B2 (en) Real-time collaborative slider-swarm with deadbands for amplified collective intelligence
US11941239B2 (en) System and method for enhanced collaborative forecasting

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17821422

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17821422

Country of ref document: EP

Kind code of ref document: A1