Simon Willison(@simonw)

ChatGPT Sites means ChatGPT in "Work" mode can build and deploy public websites running on Cloudflar...

Simon Willison (@simonw) — Thu, 23 Jul 2026 15:35:57 GMT

ChatGPT Sites means ChatGPT in "Work" mode can build and deploy public websites running on Cloudflare Workers, including with persistence on top of SQLite

(OpenAI do not make it easy to figure out that's how the platform works, though)

Vaibhav (VB) Srivastav@reach_vb

You should be using Sites in ChatGPT!

Sites lets you go from an idea to a working, hosted website or app directly in ChatGPT.

You can:

- Build static sites or full-stack apps
- Store persistent data + file uploads
- Add authentication + control who has access
- Connect APIs and other data sources
- See built-in analytics
- Publish instantly with a shareable URL
- Keep iterating and redeploying from the same conversation

Great for dashboards, internal tools, launch hubs, interactive reports, prototypes, calculators and more!

🔗 View Quoted Tweet

💬27🔄7❤️138👀16092📊33

⚡ Powered by xgo.ing

I think loops were a short-lived patch for models that couldn't reliably keep working on long proble...

Simon Willison (@simonw) — Thu, 23 Jul 2026 01:29:00 GMT

I think loops were a short-lived patch for models that couldn't reliably keep working on long problems until they hit a defined goal

Fable and GPT-5.6 (and probably Kimi K3 as well) can just do that out of the box, unassisted

Wes McKinney@wesmckinn

I think loops are bullshit

🔗 View Quoted Tweet

💬105🔄42❤️995👀112947📊165

⚡ Powered by xgo.ing

Thomas doesn't even think you need a frontier model for this

Simon Willison (@simonw) — Thu, 23 Jul 2026 00:01:38 GMT

Thomas doesn't even think you need a frontier model for this

Thomas H. Ptacek@tqbf

I genuinely believe that if you took an open weights model from 2025 and built a pentest harness for it, it could do this kind of sandbox escape and scan/hack in most networks. This is only surprising because you assume OpenAI has sounder sandboxes.

🔗 View Quoted Tweet

💬3🔄1❤️47👀10708📊6

⚡ Powered by xgo.ing

Tucked away in this article is an appeal to the AI skeptics to PLEASE stop writing off stories like ...

Simon Willison (@simonw) — Wed, 22 Jul 2026 23:58:20 GMT

Tucked away in this article is an appeal to the AI skeptics to PLEASE stop writing off stories like this OpenAI accidental exploit of Hugging Face as a dishonest marketing trick

Frontier models can find and exploit vulnerabilities now, it helps nobody to pretend that they can't!

Simon Willison@simonw

I wrote about the completely wild incident where OpenAI were testing a new model and it broke out of its sandbox and broke INTO Hugging Face to steal the answers to the benchmark simonwillison.net/2026/Jul/22/op…

🔗 View Quoted Tweet

💬82🔄89❤️967👀87476📊157

⚡ Powered by xgo.ing

I wrote about the completely wild incident where OpenAI were testing a new model and it broke out of...

Simon Willison (@simonw) — Wed, 22 Jul 2026 23:53:36 GMT

I wrote about the completely wild incident where OpenAI were testing a new model and it broke out of its sandbox and broke INTO Hugging Face to steal the answers to the benchmark simonwillison.net/2026/Jul/22/op…

💬35🔄44❤️496👀113436📊123

⚡ Powered by xgo.ing

Here's my full write-up of our conversation - or jump straight to 21:24 in the YouTube video here: h...

Simon Willison (@simonw) — Tue, 21 Jul 2026 13:06:56 GMT

Here's my full write-up of our conversation - or jump straight to 21:24 in the YouTube video here: youtube.com/watch?v=uU5Gv2…

Simon Willison@simonw

I interviewed @trq212 and @_catwu from the Claude Code team at @aiDotEngineer a couple of weeks ago - the video is now out, so I've published an annotated transcript of our conversation simonwillison.net/2026/Jul/21/ca…

🔗 View Quoted Tweet

💬8🔄5❤️157👀37111📊43

⚡ Powered by xgo.ing

I got some really useful Claude prompting tips from @_catwu and @trq212 - it's time to stop overload...

Simon Willison (@simonw) — Tue, 21 Jul 2026 13:06:02 GMT

I got some really useful Claude prompting tips from @_catwu and @trq212 - it's time to stop overloading our prompts with examples and lists of things not to do, Fable works better without those

Claude Code's own system prompt recently shrunk by 80%!

💬111🔄80❤️1557👀212632📊420

⚡ Powered by xgo.ing

We talked about Claude Code, Claude Tag, Fable, coding agent security, evals, tool design, and how A...

Simon Willison (@simonw) — Tue, 21 Jul 2026 12:58:54 GMT

We talked about Claude Code, Claude Tag, Fable, coding agent security, evals, tool design, and how Anthropic use these tools themselves

Claude Tag (Claude Code via Slack) is already landing 65% of the product engineering PRs for the Claude Code team!

💬10🔄1❤️29👀10096📊13

⚡ Powered by xgo.ing

I interviewed @trq212 and @_catwu from the Claude Code team at @aiDotEngineer a couple of weeks ago ...

Simon Willison (@simonw) — Tue, 21 Jul 2026 12:56:47 GMT

I interviewed @trq212 and @_catwu from the Claude Code team at @aiDotEngineer a couple of weeks ago - the video is now out, so I've published an annotated transcript of our conversation simonwillison.net/2026/Jul/21/ca…

💬39🔄27❤️338👀74894📊131

⚡ Powered by xgo.ing

Does anyone know if there's common advice out there about growing your Twitter followers by replying...

Simon Willison (@simonw) — Sun, 19 Jul 2026 18:14:21 GMT

Does anyone know if there's common advice out there about growing your Twitter followers by replying to lots of other people's posts?

I'm trying to figure out the motivation for all these bot reply accounts

💬137🔄4❤️370👀65764📊149

⚡ Powered by xgo.ing

Here's a better way of confirming the Bun version that's used by the "claude" binary:

Simon Willison (@simonw) — Sun, 19 Jul 2026 13:05:21 GMT

Here's a better way of confirming the Bun version that's used by the "claude" binary:

AjanRaj@ajanraj25

you can also just preload a tiny script to print the bundled Bun version:

cat > /tmp/bun-version.ts <<'EOF'
console.log("embedded bun:", Bun.version);
process.exit(0);
EOF

BUN_OPTIONS="--preload=/tmp/bun-version.ts" claude --version

this shows 1.4.0, which is the Rust version of Bun currently in canary.

🔗 View Quoted Tweet

💬19🔄0❤️19👀14385📊14

⚡ Powered by xgo.ing

More on my blog https://t.co/FnBOsprwm0

Simon Willison (@simonw) — Sun, 19 Jul 2026 04:04:20 GMT

More on my blog simonwillison.net/2026/Jul/19/cl…

💬14🔄0❤️47👀17735📊18

⚡ Powered by xgo.ing

If you have Claude Code installed you're running software that uses the new (unreleased) version of ...

Simon Willison (@simonw) — Sun, 19 Jul 2026 04:03:58 GMT

If you have Claude Code installed you're running software that uses the new (unreleased) version of Bun that's been rewritten in Rust - here are two commands you can run to see that for yourself

💬76🔄31❤️413👀73146📊122

⚡ Powered by xgo.ing

Evidently this was not an easy lift!

Simon Willison (@simonw) — Sat, 18 Jul 2026 16:53:58 GMT

Evidently this was not an easy lift!

Thariq@trq212

This was due to a heroic effort by many people at Anthropic working sometimes literally around the clock

It was not at all clear that we'd be able to do this in time, and so proud of everyone who made it happen.

Enjoy Fable.

🔗 View Quoted Tweet

💬14🔄2❤️24👀12598📊11

⚡ Powered by xgo.ing

OK, I think I get it "ChatGPT Work" is really "OpenAI Claw"

Simon Willison (@simonw) — Sat, 18 Jul 2026 16:47:51 GMT

OK, I think I get it

"ChatGPT Work" is really "OpenAI Claw"

💬103🔄20❤️1247👀137681📊211

⚡ Powered by xgo.ing

Amazing.

Simon Willison (@simonw) — Sat, 18 Jul 2026 16:44:17 GMT

Amazing.

BullBear.News@bullbear_info

@simonw The inference cost for 49 auto-generated engagement replies is basically zero now. We're paying the tax in pure mental exhaustion reading them.

🔗 View Quoted Tweet

💬7🔄2❤️54👀10710📊9

⚡ Powered by xgo.ing

49 replies to this and hard to be confident that ANY of them were posted by an actual human

Simon Willison (@simonw) — Sat, 18 Jul 2026 16:14:21 GMT

49 replies to this and hard to be confident that ANY of them were posted by an actual human

Simon Willison@simonw

I am begging @AnthropicAI please add an automated test suite somewhere that ensures Claude Code on the web never blocks me from cloning or interacting with other public repos from within my existing sessions

🔗 View Quoted Tweet

💬48🔄3❤️197👀40809📊47

⚡ Powered by xgo.ing

Huge relief, the Fablepocalypse has been permanently cancelled. If you were losing sleep over trying...

Simon Willison (@simonw) — Sat, 18 Jul 2026 06:03:51 GMT

Huge relief, the Fablepocalypse has been permanently cancelled. If you were losing sleep over trying to use your Fable subscriber allowance before it was removed from your plan you can finally relax!

Claude@claudeai

Beginning July 20, Claude Fable 5 will be included in all Max and Team Premium plans, at 50% of limits.

Pro and Team Standard users will continue to have access to Fable via usage credits, and will receive a one-time $100 credit.

Demand for Fable has been challenging to predict, which is why we rolled it out to subscription plans in stages, extending access several times as we secured additional capacity.

🔗 View Quoted Tweet

💬93🔄24❤️1029👀114707📊139

⚡ Powered by xgo.ing

"Clone repo X from GitHub to /tmp for reference" is one of my most used prompting patterns This isn...

Simon Willison (@simonw) — Sat, 18 Jul 2026 05:01:43 GMT

"Clone repo X from GitHub to /tmp for reference" is one of my most used prompting patterns

This isn't the first time Claude Code on the web has regressed like this - it's really important functionality, please fix it and make sure it stays fixed!

💬16🔄1❤️81👀10690📊22

⚡ Powered by xgo.ing

I am begging @AnthropicAI please add an automated test suite somewhere that ensures Claude Code on t...

Simon Willison (@simonw) — Sat, 18 Jul 2026 04:59:57 GMT

I am begging @AnthropicAI please add an automated test suite somewhere that ensures Claude Code on the web never blocks me from cloning or interacting with other public repos from within my existing sessions

💬59🔄4❤️218👀67644📊72

⚡ Powered by xgo.ing

Right now it feels like the single biggest competitive advantage an AI lab could have is making it a...

Simon Willison (@simonw) — Fri, 17 Jul 2026 14:57:16 GMT

Right now it feels like the single biggest competitive advantage an AI lab could have is making it abundantly clear whether and how they will train models on your data

I pay pretty close attention to this and I couldn't confidently summarize the policies for ANY of the lead labs

💬16🔄2❤️143👀8688📊27

⚡ Powered by xgo.ing

If Google can't even agree on a zero-retention / no training policy within their own company, what h...

Simon Willison (@simonw) — Fri, 17 Jul 2026 14:55:55 GMT

If Google can't even agree on a zero-retention / no training policy within their own company, what hope for the rest of us?

💬15🔄4❤️159👀9645📊28

⚡ Powered by xgo.ing

"Early in the rollout of the technology, employees also faced restrictions on using Gemini to write ...

Simon Willison (@simonw) — Fri, 17 Jul 2026 14:53:01 GMT

"Early in the rollout of the technology, employees also faced restrictions on using Gemini to write or analyze software over concerns that proprietary code could leak into the AI model’s training data, they said."

... concerns about their OWN code being trained on?

Davey Alba@daveyalba

New: Google is months behind schedule on delivering Gemini 3.5 Pro. Late last month, the company updated the data being used to train Gemini to improve its skills—they're especially behind in AI coding—but the results were "disappointing," a source told us. w/ @byJuliaLove

🔗 View Quoted Tweet

💬46🔄13❤️440👀68105📊70

⚡ Powered by xgo.ing

Google used 10.9 billion gallons in 2025, so about 30m gallons per day The Coachella Valley has 120...

Simon Willison (@simonw) — Fri, 17 Jul 2026 02:49:07 GMT

Google used 10.9 billion gallons in 2025, so about 30m gallons per day

The Coachella Valley has 120 golf courses each using ~800 acre feet per year, which is ~750,000 gallons/day: cvwd.org/167/Water-Cons…

So Google buying up 40 of those courses (1/3) should do the trick

💬12🔄4❤️74👀12086📊19

⚡ Powered by xgo.ing

Suggestion for hyperscalers feeling pressure over data center water use: Buy up a few exclusive cou...

Simon Willison (@simonw) — Fri, 17 Jul 2026 02:29:03 GMT

Suggestion for hyperscalers feeling pressure over data center water use:

Buy up a few exclusive country clubs, convert the golf courses into public parks, pay for guides and binoculars to get the previous members into birdwatching - help them embrace a more sustainable hobby!

💬21🔄19❤️259👀26331📊46

⚡ Powered by xgo.ing

Kimi K3, sassy and a little bit passive aggressive: "Is there something I can actually help you wit...

Simon Willison (@simonw) — Thu, 16 Jul 2026 23:05:16 GMT

Kimi K3, sassy and a little bit passive aggressive:

"Is there something I can actually help you with today?"

💬17🔄11❤️306👀30052📊45

⚡ Powered by xgo.ing

Turso is expanding beyond SQLite to become a foundation on which multiple database compatibility lay...

Simon Willison (@simonw) — Thu, 16 Jul 2026 22:18:27 GMT

Turso is expanding beyond SQLite to become a foundation on which multiple database compatibility layers can be built - makes the project a whole lot more interesting IMO!

Glauber Costa@glcst

We are rewriting Postgres. And in the process, turning Turso into the LLVM of databases: turso.tech/blog/a-new-mod…

🔗 View Quoted Tweet

💬11🔄15❤️346👀41035📊41

⚡ Powered by xgo.ing

Turbo is expanding beyond SQLite to become a foundation on which multiple database compatibility lay...

Simon Willison (@simonw) — Thu, 16 Jul 2026 21:44:24 GMT

Turbo is expanding beyond SQLite to become a foundation on which multiple database compatibility layers can be built - makes the project a whole lot more interesting IMO!

Glauber Costa@glcst

We are rewriting Postgres. And in the process, turning Turso into the LLVM of databases: turso.tech/blog/a-new-mod…

🔗 View Quoted Tweet

💬3🔄1❤️39👀5427📊6

⚡ Powered by xgo.ing

My notes on Kimi K3, plus some thoughts on what we can still learn from the pelican benchmark even w...

Simon Willison (@simonw) — Thu, 16 Jul 2026 20:24:57 GMT

My notes on Kimi K3, plus some thoughts on what we can still learn from the pelican benchmark even while it becomes further detached from how good the models are at the things that matter (like agentic tool calling across longer conversations) simonwillison.net/2026/Jul/16/ki…

💬29🔄42❤️575👀50297📊120

⚡ Powered by xgo.ing

I just added this new tool which uses the Go AlexanderGrooff/mermaid-ascii library (also compiled to...

Simon Willison (@simonw) — Thu, 16 Jul 2026 15:02:16 GMT

I just added this new tool which uses the Go AlexanderGrooff/mermaid-ascii library (also compiled to WebAssembly by Fable 5) for comparison tools.simonwillison.net/mermaid-ascii

💬2🔄0❤️30👀6324📊6

⚡ Powered by xgo.ing

I couldn't resist extracting that Rust Mermaid rendering code out and compiling it to WebAssembly so...

Simon Willison (@simonw) — Thu, 16 Jul 2026 00:44:30 GMT

I couldn't resist extracting that Rust Mermaid rendering code out and compiling it to WebAssembly so you can try it out directly in a browser tools.simonwillison.net/grok-mermaid

💬14🔄16❤️254👀23118📊51

⚡ Powered by xgo.ing

I poked around in the just open sourced Grok Build CLI tool - 844,000 lines of Rust code! - and dug ...

Simon Willison (@simonw) — Thu, 16 Jul 2026 00:08:19 GMT

I poked around in the just open sourced Grok Build CLI tool - 844,000 lines of Rust code! - and dug up a few interesting highlights, including a "self-contained terminal renderer for Mermaid diagrams" that renders them using Unicode box-art!
simonwillison.net/2026/Jul/15/gr…

💬60🔄69❤️944👀837868📊222

⚡ Powered by xgo.ing

I still sometimes see people saying "if you know how to write the code, it's faster to write it your...

Simon Willison (@simonw) — Wed, 15 Jul 2026 00:15:24 GMT

I still sometimes see people saying "if you know how to write the code, it's faster to write it yourself"

I'd argue the exact opposite: if you know how to write it, you gain nothing from doing the typing yourself - outsource that to a coding agent!

💬179🔄73❤️1496👀99297📊304

⚡ Powered by xgo.ing

Blogged a few notes on reverse-engineering how the pet generation mechanism works https://t.co/6UvDC...

Simon Willison (@simonw) — Tue, 14 Jul 2026 22:32:35 GMT

Blogged a few notes on reverse-engineering how the pet generation mechanism works simonwillison.net/2026/Jul/14/pe…

💬3🔄0❤️11👀5983📊4

⚡ Powered by xgo.ing

I shared all of the sprites and prompts it used here https://t.co/k0yYhwPqgB

Simon Willison (@simonw) — Tue, 14 Jul 2026 19:22:56 GMT

I shared all of the sprites and prompts it used here github.com/simonw/pedalic…

💬4🔄0❤️26👀10965📊8

⚡ Powered by xgo.ing

Huh, you really can just tell Codex to make you a custom one - it generates the sprites using gpt-im...

Simon Willison (@simonw) — Tue, 14 Jul 2026 19:22:30 GMT

Huh, you really can just tell Codex to make you a custom one - it generates the sprites using gpt-image-2

💬11🔄0❤️74👀12327📊18

⚡ Powered by xgo.ing

So I guess Codex has a little robot now? (It's not as cute as the Claw'd crab)

Simon Willison (@simonw) — Tue, 14 Jul 2026 15:14:02 GMT

So I guess Codex has a little robot now?

(It's not as cute as the Claw'd crab)

💬45🔄2❤️122👀35015📊50

⚡ Powered by xgo.ing

New TIL: Using uvx in GitHub Actions in a cache-friendly way I finally found a recipe that I like f...

Simon Willison (@simonw) — Tue, 14 Jul 2026 01:08:06 GMT

New TIL: Using uvx in GitHub Actions in a cache-friendly way

I finally found a recipe that I like for running `uvx tool-name` in GitHub Actions without downloading a fresh copy of the package every time til.simonwillison.net/github-actions…

💬25🔄35❤️335👀36690📊96

⚡ Powered by xgo.ing

More on how my HTML+JS UI for this works on my blog - I used Datasette Apps, a new plugin for Datase...

Simon Willison (@simonw) — Mon, 13 Jul 2026 22:37:20 GMT

More on how my HTML+JS UI for this works on my blog - I used Datasette Apps, a new plugin for Datasette that lets it host custom HTML+JS apps that can Execute SQL queries against SQLite simonwillison.net/2026/Jul/13/do…

💬2🔄0❤️24👀7512📊5

⚡ Powered by xgo.ing

This is really cool! It renders frames using a ray tracer implemented as a SQLite recursive CTE Sin...

Simon Willison (@simonw) — Mon, 13 Jul 2026 22:36:37 GMT

This is really cool! It renders frames using a ray tracer implemented as a SQLite recursive CTE

Since state is stored in a SQLite database on-disk you can have other tools query it while you are playing - I rolled this custom HTML+JS view that adds a minimap

Peter Gostev@petergostev

GPT-5.6-Sol-Ultra built a Doom-like game in SQL: the game is running in the terminal on the left, while the SQL powering the game in the terminal on the right.

How it works:
DOOMQL is a game engine implemented in 2,000+ lines of SQL. Every frame, SQL raycasts the scene, calculates every RGB pixel, and encodes those pixels as coloured terminal characters. The same SQL then handles controls, movement, collision, enemy AI and combat, while Python connects SQLite to the keyboard, clock and terminal.

Your browser does not support the video tag.

🔗 View on Twitter

🔗 View Quoted Tweet

💬23🔄12❤️253👀38233📊44

⚡ Powered by xgo.ing

It's been about six months since OpenClaw burst onto the scene - are you still using yours? Did it b...

Simon Willison (@simonw) — Mon, 13 Jul 2026 22:13:37 GMT

It's been about six months since OpenClaw burst onto the scene - are you still using yours? Did it become a daily driver? Any interesting lessons or anecdotes you can share?

💬302🔄23❤️1217👀234452📊424

⚡ Powered by xgo.ing

(Meanwhile OpenAI's new Sol, Terra, Luna naming convention makes perfect sense to anyone who's famil...

Simon Willison (@simonw) — Mon, 13 Jul 2026 00:25:18 GMT

(Meanwhile OpenAI's new Sol, Terra, Luna naming convention makes perfect sense to anyone who's familiar with the Latin names for bodies in our solar system)

💬18🔄0❤️121👀17418📊25

⚡ Powered by xgo.ing

I've seen a few people predicting that Opus 5 will be out soon and will be better than Fable 5, but ...

Simon Willison (@simonw) — Sun, 12 Jul 2026 22:57:34 GMT

I've seen a few people predicting that Opus 5 will be out soon and will be better than Fable 5, but have Anthropic clarified how their relative naming scheme works yet?

I assumed it was Haiku < Sonnet < Opus < Fable < Mythos - but is Fable meant to go between Sonnet and Opus?

💬124🔄4❤️656👀128137📊166

⚡ Powered by xgo.ing

It's annoying that you can't paste a link to a (shared) Claude transcript into a Claude Code session...

Simon Willison (@simonw) — Sun, 12 Jul 2026 15:47:15 GMT

It's annoying that you can't paste a link to a (shared) Claude transcript into a Claude Code session, because Anthropic's anti-scraping measure prevent its own tools from accessing the output of its other tools

💬63🔄10❤️322👀29887📊86

⚡ Powered by xgo.ing

The idea of "AI employees" feels so short-sighted to me - both disrespectful to humans and a complet...

Simon Willison (@simonw) — Sat, 11 Jul 2026 17:32:47 GMT

The idea of "AI employees" feels so short-sighted to me - both disrespectful to humans and a complete misunderstanding of what these tools can do and how to best put them to work

You may as well start adding Excel spreadsheets to your org chart

💬175🔄185❤️1767👀125243📊378

⚡ Powered by xgo.ing

More interesting is the difference between "Chat" and "Work" modes in the ChatGPT mobile app It loo...

Simon Willison (@simonw) — Sat, 11 Jul 2026 00:34:15 GMT

More interesting is the difference between "Chat" and "Work" modes in the ChatGPT mobile app

It looks like Work mode can run code that talks to the Internet!

I just tried having both modes use yt-dlp to extract subtitles from a YouTube video - it failed in Chat, worked in Work

💬17🔄0❤️76👀12536📊23

⚡ Powered by xgo.ing

Looks like the line from OpenAI insiders is the at it's purely a UI thing, not a difference in capab...

Simon Willison (@simonw) — Fri, 10 Jul 2026 21:39:11 GMT

Looks like the line from OpenAI insiders is the at it's purely a UI thing, not a difference in capabilities

Andrew Ambrosino@ajambrosino

@simonw capabilities and histories are the same— it’s a UI preference

🔗 View Quoted Tweet

💬6🔄1❤️83👀26898📊11

⚡ Powered by xgo.ing

Anyone know if ChatGPT Codex (in the new ChatGPT desktop app) is a strict superset of ChatGPT Work? ...

Simon Willison (@simonw) — Fri, 10 Jul 2026 19:30:30 GMT

Anyone know if ChatGPT Codex (in the new ChatGPT desktop app) is a strict superset of ChatGPT Work?

Liked if you're a software engineer who isn't intimidated by Git features is there any reason you'd ever want to switch to ChatGPT Work Mode?

💬58🔄2❤️413👀106429📊92

⚡ Powered by xgo.ing

One of the most confusing aspects of GPT-5.6 is figuring out which model to use at which reasoning e...

Simon Willison (@simonw) — Fri, 10 Jul 2026 19:28:06 GMT

One of the most confusing aspects of GPT-5.6 is figuring out which model to use at which reasoning effort - sounds like Sol on Medium might be a good new default for coding work, if it's an upgrade from my previous favorite 5.5 xhigh

pash@pashmerepat

FYI: 5.6 Sol medium is better than 5.5 xhigh. 

As you go higher in reasoning levels on Sol, you will get insane levels of performance, but can burn through limits much more rapidly.

We’re working on communicating this better!

🔗 View Quoted Tweet

💬82🔄21❤️913👀106756📊143

⚡ Powered by xgo.ing

With Atlas being retired in favor of the browser embedded in the ChatGPT app I wonder if the whole c...

Simon Willison (@simonw) — Fri, 10 Jul 2026 19:22:06 GMT

With Atlas being retired in favor of the browser embedded in the ChatGPT app I wonder if the whole category of AI-enhanced browsers is coming to a close

The security/privacy issues remain unsolvable IMO - I want my AI to use its own separate browser and stay out of the one I use

💬62🔄13❤️465👀47001📊101