i'm going dark on socials to try and ship an mvp in 7 days 👋 playbooks 3.0: the final (final) dance

@kevinyun Lovable eh?

when x sneakily switches you back to the for you tab

@pvncher @abacaj hey I just downloaded repoprompt- what’s the best guide on how to get the best out of it?

@MilaChervenkova yup! I’m 20 years into that search 😂

been thinking about this tweet a lot recently in 2024 I had exactly 0x agency (for valid reasons, but still) this year is the time to push through and rediscover the drive of the past…

@jonkkillian @togethercompute @groq I have no clue! says $0 so far... 🤞

@ED84VG your best bet is to actually use them both and see how it feels for you in your tasks. i think o1 is prob better but the cost difference is ridiculous

@jonkkillian whats the best US hosted api you've found?

@jonkkillian @GroqInc yeah, and fine for what i'm working on!

what’s the best mac app to chat with any llm?

@dannypostmaa haha it’s what some call the new 3.5 nothing to see here 😂

o3-mini-high-new-final-final shipping soon..

@mattshumer_ Would you still still use a model like o1 to create implementation details? Then pass to o3 to code it?

@delicto_nick @asaio87 i've always published them at once (sometimes millions), but i'm seeing good results by limiting my scope and focus to a much smaller set of 100-1000 pages at a time for 30k I might try in smaller batches!

@AdamFx the ai is used to take the page html and turn it into structured data obviously could be done without ai but that means the user figuring out the dom elements to scrape, which is an unneeded bottleneck when ai calls are so cheap!

@SemanticEntity been busy haha! 👀