Joe Dwyer

hey @Uber why is it so hard to report security issues in your @UberEats app? Attempted hijack of my account, with full order and phone number (meaning someone at the restaurant / your team was complicit). Scammer ended up threatening violence when I called him out.

People are misreading the SpaceX/Cursor deal as an M&A story. It’s actually a bet on what the real bottleneck in frontier coding models is. xAI has struggled to close the gap with Claude Code and Codex. Cursor sits on the best corpus of developer traces in the world. The deal lets Cursor train Composer on Colossus while xAI runs the same recipe on Grok. Both sides find out, at the same time, whether Cursor's data is actually the difference. The option structure reflects that uncertainty. If the training work ports over, SpaceX buys Cursor and owns the pipeline. If it doesn’t, they pay $10B for the experiment and walk. Either outcome, Grok ends up stronger than it would have been, and xAI gets an answer to a question it couldn’t answer internally. The part worth holding onto: a pre-IPO company just priced a live experiment to figure out whether real developer traces are the scarce input in coding agents. $10B is what they’re paying to run it. $60B is what the answer is worth if it comes back yes.

Mythos is not a bad name for a model but it would be better if Anthropic switched to using famous Claudes. Monet, Debussy etc. The final model that achieves AGI would obviously be Van Damme

Friends outside of tech: lol copilot is dumb Friends in tech: I just bought iodine tablets and have made an offer on land upstate. My supplies of antibiotics and potable water are sufficient but I need to set up hydroponics to make it through the first few years.

@Bquittem @LynAldenContact I really enjoyed it as well. Great story / characters / pacing, all played out in a very interesting sci-fi context.

Update: finished @LynAldenContact book "The Stolguard Incident" 9/10 sci-fi, highly recommended. When screenplay?

NEW DETAILS 🚨: NASA astronaut Mike Fincke who suffered a mysterious ‘medical emergency’ on the ISS earlier this year was prepping for a spacewalk when he suddenly lost the ability to speak It lasted about 20 minutes and then he was completely fine Doctors have ruled out a stroke/heart attack but still have no explanation

Just experienced the first instance where an agent I built (with AI help) diagnosed a problem with its own code and suggested a fix. Hold on to your hats people.

My take on California’s “One-Time” Billionaire Tax. It’s much worse than it looks. 📉 Will it pass? Yes, likely. It only needs 50%+1 voter approval. SEIU + CTA have done this before—Prop 55 won 63% in 2016. ⚖️ Will it get tied up in litigation? Almost certainly. Retroactive wealth tax on a tax type CA has never had = due process challenges. Billionaires have the legal budgets for years of fights. 👋 But it’s clearly … only the start. The goal is an annual tax, not one-time. And the target is $25m-50m net worth folks, including illiquid foldings (early stage founders raising a Series B). The “one-time” framing is strategic, not terminal. The same coalition (CTA, CFT, SEIU) already has AB 259 written—an >annual< 1% wealth tax at a >$50M threshold<, with plans to go to $25M. It’s been introduced 3 years running. The one-time tax removes the constitutional barrier. Once that’s gone, the annual version becomes a much easier ballot measure. 🔃 CA Policy Center said: “If SEIU hopes to keep Medi-Cal spending growing, it may need to place repeated wealth taxes on the ballot—potentially lowering the threshold as billionaires flee.” The real risk for founders: At $1B, you’re taxing ~200 people. At $50M, you’re taxing 23,000 households—including most successful founders on paper before any liquidity event. The rational move isn’t to leave when you hit the threshold. It’s to leave—or never incorporate in CA—before you get anywhere close. ✈️ Net net: it will make sense to leave before the Series B. Vinod Khosla nailed it: “Even people who don’t expect this initiative to pass are still planning to leave because there will be another one.” The one-time tax is the constitutional Trojan horse. The annual tax — at a much lower threshold — is already waiting inside.

For Christmas I want a printout of all the posts that have shown up right as I open X that disappear forever after half a second when the page refreshes

Might need to include “massive” and “insane” too

Is there a way to filter out all posts that include “literally?” (“This guy literally built a ….”). Almost always useless content. And almost always an improper use of the word.

My life got a lot better when I realized that small talk is just a TCP handshake for humans Skipping either will result in bad data transfer

@Matt_Pinner Mildred (“Mimi”)

WHAT was your grandmother's name... I'm looking for a vintage girl's name. Not Betty!

@cursor_ai How do prompt this to get it to work? I've got it all enabled, but my agent can't (or won't) see the page content, take screenshots, etc. Is there documentation for this somewhere I didn't see?

Cursor can now control your browser. Agent can take screenshots, improve UI, and debug client issues. Try our early preview with Sonnet 4.5.

@jessethanley This is insane

PSA: Declare an incident if someone on your team installed the postmark-mcp on their machine. All your emails had a secret BCC added to them since version 16.

@fortelabs This really resonated with me

Best succinct description of the increasingly widening difference between ChatGPT and Claude I've seen recently Source below

@SinaHartung That’s cool.

it has come to my attention that this is not universal knowledge you can just type docs.new or sheets.new into your browser and it will immediately open a new google doc or sheet

@Scobleizer @sama @IllusionOfLife @kevinafischer This was super interesting and thought provoking. Thank you for the work you do.

The device that @sama and Jony Ive is building. I lay out why it will make Apple far less relevant in our lives. And I didn't even go into the medical impact of this. Thanks @IllusionOfLife and @kevinafischer, among many others working in augmented reality and robotics, for the inspiration.

This paper finds LLMs' ability to understand that others have different beliefs (Theory of Mind) comes from 0.001% of their parameters. Break those specific weights & the model loses both its ability to track what others know AND language comprehension. Interesting implications.

@alessio_joseph This is an engineering marvel that you have wrought. Still hard to believe it can be so.

i just got fired from apple. my job was to make it so finder gives you zero results even when you search exact name matches and file formats