Sup*

Pause.. Maybe?

@shashankgoyal95 @rauchg The 403s

@mostlyaboutai @rauchg Why did you need to get off of OR?

Bet on every model at once vercel.com/ai-gateway

@rauchg @tomocchino x2

gm 🇦🇷

@sundeep All of a sudden open weights are interesting again 👀

@Kimi_Moonshot is killing it! Web based 3D has never been this good!

@shadcn Huge thanks for everything you build, Shadcn ftw!

Thank you to my new sponsors.

@elliotarledge :) Sheets off

im not gonna sleep well tonight

@waterloo_intern @theo fwiw, I expanded on the benchmark saturation point here: x.com/mostlyaboutai/…

@waterloo_intern These are all very valid points. Benchmarks are now almost useless the IRL application is still something that matters, and imo, @theo brings out that part of the models really well. Also, it's funny how os community & models are no longer limited to China.

theo, love your takes, but i have to disagree: a) the cost here is as it stands today. before inference providers get the weights. before performance optimizations. b) for larger orgs, they get dedicated NODES of the model. they don't pay $ per token, they pay $ per gpu hour. this is HUGE. they get their own endpoint. the $ savings are huge. the model is never down (like, ever.) c) you OWN the guardrails. not at mercy of a lab telling you what you can / can't do with the tokens you pay for. d) most importantly, you'll never have to worry about closed-source pulling something funny behind scenes / making the model dumber the weights are right there. if they change, people can see. they won't need to speculate ... (which, as your own catalog shows, is a full-time job)

@elliotarledge 👀

should i make this my wallpaper?

Agreed But, I wouldn't call them superintelligence as that's a form of AI that's wayy superior in every aspect. We are more like on MajorIntelligence/Pre-AGI progressing towards AGI. 1. I agree about people becoming irrelevant by losing their personality. And ironically, that's what is happening right now. People are losing personality while AI that isn't supposed to have one is becoming customizable to its user's liking (so much so, that as a result humans tend to share all their personal and emotional vulnerabilities with AI). 2. AI isn't novel so unless it is taught to write in a certain fashion it will write in the way it is trained by data.

The term "AGI" has aged very poorly. AI can't possibly me more different from human intelligence. It's *far better* than human intelligence, for most economically-relevant tasks. But that doesn't make AIs better than humanity. There are wonderful things humans can do that these superintelligences (the better term) can't. For one, we care-for and care-about other humans in a powerful and inherent way, that we then teach machines to emulate. We must keep that as the top priority in everything we do. We have to be as pro-human-life as it gets. AIs can replace tasks you do, but they can't replace the proverbial you. This is why they horribly suck at writing! Even when given a corpus of your writing and 200 Skills and 50 subagents running in gRaPh LoOps, they produce robotic prose… because they are robots. I know one definitive way that people can become irrelevant. They stop being themselves: lose their identity, delegate all their writing, their unique thoughts, the creative ways in which they can use the machines. Opting out of 'weighing in'. I think we're safe though. There's nothing more distasteful right now than AI replies in social media or (🟢 T H I S T H I N G ) in landing pages. Quality and humanity will prevail.

@nikitabier "Encara Nikita" You looked a lot like Messi

This is who runs this account

Hey @kuizinas Can we have a 'Takeover' mode in Contra? For times when I'd like to take over the chat from the co-agent. Rn, the Co-agent just keeps going and I have no specific in-chat mode to pause it. Contra mobile :)

Shhhh...

If anybody wants to send me freebies but is too shy to ask, my DMs are open. No need to hesitate. 👀

@TheRealAdamG That's what I am building Curative Intelligence for ⚡ Good to see that I am on the right track

openai.com/index/a-scorec… “The ultimate scorecard for the age of AI could be looked at as ‘Useful Intelligence per Dollar.’ This metric answers four key questions: 1. Is AI completing work that matters? 2. What does each successful task cost? 3. Can people depend on the result? 4. Does each AI dollar produce more value as usage grows?”

@TheRealAdamG Tweets like these show progress ⚡👀

GPT-5.6 is even more thorough and dependable.