🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

only a matter of time before there's a gpt4o generated picture in an art museum

don't search, think from first principles

I'm crowdsourcing the most useful AI use cases, infra & products in 2025. Top AI founders, researchers & VCs are sharing their AI stacks. Contribute through the form in my bio and I'll: * email you the results * give you a shoutout Comment "alpha" below and I'll DM you results.

the IDE war begins

only FAF can host a demo day the week before finals and get insane turnout + the most incredible vibes from our epic community 🚀😮‍💨 here's what stanford's builders have been cooking this quarter: 🧵

just noticed Cursor has sound effects when coding (when errors appear, code runs successfully, terminal commands run) does this mean a SWE makes music when coding well? and background agents would make music like a band or orchestra? with the human listening as the conductor

I'm hosting a small community where AI power users (who spend $500+/mo on AI) share learnings while optimizing their AI stacks and workflows. Reply w/ how much $ you spend on AI/month to request an invite.

see the AI demos Stanford student built last quarter👇👇

love this blog post about ai x gaming. it's maybe the most undervalued application for AI. also kevin's blog is 10/10- i kind of want to keep it a secret

So I think something else that doesn't get discussed much is the extrapolation of this inference : training trend - 2015: back in the day, we would train one model per dataset, and inference it once (to obtain the eval result for our paper) - 2020: with chatgpt, multi-task

(Tegus+Accenture).ai

in next year or so, all the cool kids are going to have mac minis / open AI agent boxes plugged in at home

Why you should stop working on RL research and instead work on product // The technology that unlocked the big scaling shift in AI is the internet, not transformers I think it's well known that data is the most important thing in AI, and also that researchers choose not to work