Programmer + founder currently building https://xitter.cf/t.co/ZSAJ7QVz8k Previously worked in evals/benchmarks, improving AI models at the the top AI research labs

San Francisco, CA
Joined August 2026
Include

Only show posts containing:

Exclude

Hide posts containing:

Time range
-
Minimum likes
Replying to @AleksDoesCode
Nice! Will give it a try! Where are you from btw? Serbian? Haha
1
1
51
Replying to @AleksDoesCode
Would you recommend having your agent apply them gradually? For example a few submissions each day, so it isn’t sketchy? Also does your “skill” or however you set it up, have a way where it can present to you all the different submissions it has ready and then you can approve/reject?
1
1
760
Replying to @themacliu
Coffee- I’m also in SF, would be cool to meet up
43
Replying to @OpenAIDevs
It was only a matter of time for this to release. The question is how does the pricing differ from just running codex the old fashioned way?
389
Replying to @morellifit
Lmao what kind of study or basis is this off of? Fentanyl is like like dead last but I’m pretty sure anyone addicted to it in the tenderloin would certainly tell you it it’s definitely life changing, just not for the better
2
25
7,682
Replying to @steipete @ehzish
Are you referring to slop as too much code without tests or the other way around - too many tests for the code? I’m confused now haha. In general you go for about 2X test code as prod code?
80
Replying to @VadimStrizheus
Maybe because they’re competitors
1
1,325
Replying to @mitsuhiko
Just reset man, get out of that negative / non productive loop
194
Replying to @ZachWarunek
Sure can you review this first? Thanks
2
236
Replying to @signulll
That pretty much is what they do already. Have you seen the openAI billboards in SF for example?
103
Replying to @Suhail
Anyone remember when we used to write code by hand and you put up a PR and get a comment on how you can reduce a 100 line code snippet down to 90 lines instead? I wonder where that energy went
16
Replying to @Joey_Walker82
Same, good luck 💪
1
6
Replying to @elonmusk
Anyone have a list or community of legit technical / AI follows? All I get is AI articles with thousands of likes and bookmarks and you actually read the article and it’s total garbage
5
Replying to @ThePrimeagen
Blame it on the benchmarks and the way that everyone has measured “the best model” up until now. There definitely seems like a vibe shift happening where people are realizing that more intelligence isn’t the only thing that matters tho
432
Replying to @kirtandopamine
This was anthropics specialty but the vibe has completely shifted in the last couple months… it really has to be studied haha
1
15
Replying to @richardhutton
Are you just running every prompt continuously within the same thread or something? This is near impossible unless something is wrong
1
1,613
Replying to @a16z
This guy has no idea what he’s talking about and I haven’t even bothered watching the video
77
Replying to @natiakourdadze
Can you provide the links to the business insider (or any other) of those publications?
1
1
116
Basic ass question that doesn’t have any correct answer yet gets more replies than anything else here on tech X
Your SaaS is hitting millions of requests per day Which database are you betting on? A. PostgreSQL B. MySQL C. MongoDB D. DynamoDB
2
62