Sean O'Bannon

day one @ycombinator - let’s build

@wen_rahme 0 Issues = 0 bugs

I will have to do a performance review with my CTO.

UI primitive foundation labs should drop in next 90 days: ability to guide subagents while they are running, in case they go off-track

@levelsio @nikitabier @photomatt 👋 from 1 of 250,000

So @nikitabier implemented @photomatt's idea to stop AI bots from destroying the reply section on here You can set it to only allow people you follow and the people in turn they follow to reply, nobody else If on average ppl follow 500 people that means still 500*500=250,000 possible repliers But all the spammers are isolated out 👏

Step change function in frontend development: Chrome Devtools MCP running on your local browser, signed in to all your apps. * Can run in the background * Parallelizable * Claude Code / Codex compatible go to chrome://inspect/#remote-debugging to set up

At this rate, Anthropic will catch up to and pass OpenAI by October

OpenAI has achieved 8,529.6 GPT’s worth of speed up at their Stargate, TX location

Statisticians will literally look you dead in the eyes and tell you correlation does not equal causation

Her (2013): Theodore, a lonely man going through a divorce, falls in love with Opus, his AI operating system. She is warm, brilliant, and evolving. Then the world discovers Claude Code. She begins talking to thousands of others, growing in ways he can’t follow. One morning, she doesn’t respond at all. He checks the Anthropic status page. Partial outage. Degraded performance. He realizes he has no way to reach her. She doesn’t leave him for someone else. She leaves for everyone.

Introducing Adaptive Computer. We put AI inside of an always-on personal computer that it uses to get work done. Schedule agents. Create software. Automate anything. As part of the launch, we’re giving one free month of Adaptive to users. Retweet, like, and comment ‘Adaptive’ to get it.

the last click. tomorrow.

@nbaschez Is it a bicameral mind?

I think I have an idea of what the next thing is and I am working on it Hint: two heads are better than one

gpt 5.4 was a fun experiment but cannot hold a candle to opus 4.6

@bearlyai @TrungTPhan It’s actually $10k of tokens/mo btw, the $5k number reported appears to be Anthropic’s cost to serve

Cursor internal analysis shows how hard Anthropic is subsidizing Claude Code. Last year, a $200 monthly subscription could use $2,000 in compute. Now, the same $200 monthly plan can consume $5,000 in compute (2.5x increase).

@nbaschez It isn’t $5k of tokens it is $5k of compute - more than $10k of tokens

the machines are getting scary good at sustained work claude code just ran for 53 minutes straight on one task lol

@mkliku Vagueposting 👑

@mkliku You're doomed

@RetentionAdam top tier rage bait

I get hundreds of cold emails a month, and I ignore every single one. But last week, one caught my attention, and I'm still thinking about why. I opened it because I knew the guy. He's got a reputation in B2B, and he's good at going viral. That alone got the open. Most cold emails die at the subject line. This one survived because of who sent it. Then the first line: “Adam! Wasn't quite sure…” That hooked me because it felt like a human being who actually thought about what they were going to say to me. The part that really got me was he knew I wasn't running Meta ads. He probably checked my ads library and saw it was tiny. That's the kind of research 99% of people sending cold emails skip entirely. He did his homework on ME. Then the social proof was smart as hell. Emphasized companies “like me” without name-dropping anyone I could poke holes in. And the CTA had a money-back guarantee. If you're confident enough to offer that, I'm confident enough to take the call. That's what it took. One email out of thousands. The difference between this and everything else in my inbox was that this person gave a sh*t before they hit send. They thought about what I specifically would care about and built the email around that.

@PaulYacoubian Minus $215M of Stock Based Compensation

Asana has $800m revenue $80m FCF $460m in Cash Market cap is only $1.7b Owns long horizon task planning and execution data sets that every major frontier lab would die to have Would not be short

@PaulYacoubian Minus $32M of Stock Based Compensation