Simon Willison’s Weblog

The most popular blogs of Hacker News in 2025 (via) Michael Lynch maintains HN Popularity Contest, a site that tracks personal blogs on Hacker News and scores them based on how well they perform on that platform.

The engine behind the project is the domain-meta.csv CSV on GiHub, a hand-curated list of known personal blogs with author and bio and tag metadata, which Michael uses to separate out personal blog posts from other types of content.

I came top of the rankings in 2023, 2024 and 2025 but I'm listed in third place for all time behind Paul Graham and Brian Krebs.

I dug around in the browser inspector and was delighted to find that the data powering the site is served with open CORS headers, which means you can easily explore it with external services like Datasette Lite.

Here's a convoluted window function query Claude Opus 4.5 wrote for me which, for a given domain, shows where that domain ranked for each year since it first appeared in the dataset:

with yearly_scores as (
  select 
    domain,
    strftime('%Y', date) as year,
    sum(score) as total_score,
    count(distinct date) as days_mentioned
  from "hn-data"
  group by domain, strftime('%Y', date)
),
ranked as (
  select 
    domain,
    year,
    total_score,
    days_mentioned,
    rank() over (partition by year order by total_score desc) as rank
  from yearly_scores
)
select 
  r.year,
  r.total_score,
  r.rank,
  r.days_mentioned
from ranked r
where r.domain = :domain
  and r.year >= (
    select min(strftime('%Y', date)) 
    from "hn-data"
    where domain = :domain
  )
order by r.year desc

(I just noticed that the last and r.year >= ( clause isn't actually needed here.)

My simonwillison.net results show me ranked 3rd in 2022, 30th in 2021 and 85th back in 2007 - though I expect there are many personal blogs from that year which haven't yet been manually added to Michael's list.

Also useful is that every domain gets its own CORS-enabled CSV file with details of the actual Hacker News submitted from that domain, e.g. https://hn-popularity.cdn.refactoringenglish.com/domains/simonwillison.net.csv. Here's that one in Datasette Lite.

# 7:10 pm / hacker-news, sql, sqlite, datasette, datasette-lite, cors

Was Daft Punk Having a Laugh When They Chose the Tempo of Harder, Better, Faster, Stronger? (via) Depending on how you measure it, the tempo of Harder, Better, Faster, Stronger appears to be 123.45 beats per minute.

This is one of those things that's so cool I'm just going to accept it as true.

(I only today learned from the Hacker News comments that Veridis Quo is "Very Disco", and if you flip the order of those words you get Discovery, the name of the album.)

# 5:57 am / music

Oxide and Friends Predictions 2026, today at 4pm PT (via) I joined the Oxide and Friends podcast last year to predict the next 1, 3 and 6 years(!) of AI developments. With hindsight I did very badly, but they're inviting me back again anyway to have another go.

We will be recording live today at 4pm Pacific on their Discord - you can join that here, and the podcast version will go out shortly afterwards.

I'll be recording at their office in Emeryville and then heading to the Crucible to learn how to make neon signs.

# 4:53 pm / podcasts, ai, llms, oxide

It’s hard to justify Tahoe icons (via) Devastating critique of the new menu icons in macOS Tahoe by Nikita Prokopov, who starts by quoting the 1992 Apple HIG rule to not "overload the user with complex icons" and then provides comprehensive evidence of Tahoe doing exactly that.

In my opinion, Apple took on an impossible task: to add an icon to every menu item. There are just not enough good metaphors to do something like that.

But even if there were, the premise itself is questionable: if everything has an icon, it doesn’t mean users will find what they are looking for faster.

And even if the premise was solid, I still wish I could say: they did the best they could, given the goal. But that’s not true either: they did a poor job consistently applying the metaphors and designing the icons themselves.

# 7:30 pm / apple, design, macos, usability

A field guide to sandboxes for AI (via) This guide to the current sandboxing landscape by Luis Cardoso is comprehensive, dense and absolutely fantastic.

He starts by differentiating between containers (which share the host kernel), microVMs (their own guest kernel behind hardwae virtualization), gVisor userspace kernels and WebAssembly/isolates that constrain everything within a runtime.

The piece then dives deep into terminology, approaches and the landscape of existing tools.

I think using the right sandboxes to safely run untrusted code is one of the most important problems to solve in 2026. This guide is an invaluable starting point.

# 10:38 pm / sandboxing, ai, generative-ai, llms

How Google Got Its Groove Back and Edged Ahead of OpenAI (via) I picked up a few interesting tidbits from this Wall Street Journal piece on Google's recent hard won success with Gemini.

Here's the origin of the name "Nano Banana":

Naina Raisinghani, known inside Google for working late into the night, needed a name for the new tool to complete the upload. It was 2:30 a.m., though, and nobody was around. So she just made one up, a mashup of two nicknames friends had given her: Nano Banana.

The WSJ credit OpenAI's Daniel Selsam with un-retiring Sergei Brin:

Around that time, Google co-founder Sergey Brin, who had recently retired, was at a party chatting with a researcher from OpenAI named Daniel Selsam, according to people familiar with the conversation. Why, Selsam asked him, wasn’t he working full time on AI. Hadn’t the launch of ChatGPT captured his imagination as a computer scientist?

ChatGPT was on its way to becoming a household name in AI chatbots, while Google was still fumbling to get its product off the ground. Brin decided Selsam had a point and returned to work.

And we get some rare concrete user numbers:

By October, Gemini had more than 650 million monthly users, up from 450 million in July.

The LLM usage number I see cited most often is OpenAI's 800 million weekly active users for ChatGPT. That's from October 6th at OpenAI DevDay so it's comparable to these Gemini numbers, albeit not directly since it's weekly rather than monthly actives.

I'm also never sure what counts as a "Gemini user" - does interacting via Google Docs or Gmail count or do you need to be using a Gemini chat interface directly?

Update 17th January 2025: @LunixA380 pointed out that this 650m user figure comes from the Alphabet 2025 Q3 earnings report which says this (emphasis mine):

"Alphabet had a terrific quarter, with double-digit growth across every major part of our business. We delivered our first-ever $100 billion quarter," said Sundar Pichai, CEO of Alphabet and Google.

"[...] In addition to topping leaderboards, our first party models, like Gemini, now process 7 billion tokens per minute, via direct API use by our customers. The Gemini App now has over 650 million monthly active users.

Presumably the "Gemini App" encompasses the Android and iPhone apps as well as direct visits to gemini.google.com - that seems to be the indication from Google's November 18th blog post that also mentioned the 650m number.

# 3:32 pm / google, ai, openai, generative-ai, llms, gemini, nano-banana