Pan Anon

@PalmerLuckey @JoshLeibner Don't bother engaging with morons. It's not good for you.

@JoshLeibner Hi Josh, I understand it is tough being an idiot, but opposing taxpayer-funded government programs is not pro-regulation. If it was, the ultimate anti-regulation position would be to advocate for more government programs. But that is not the case, because we aren't stupid.

Palmer Luckey: "I literally said I am anti-regulation. This isn't hard to understand." Also Palmer Luckey:

@tszzl Is it recognizable and full of tics because of watermarks, like gpt-image-2?

the frontier models tend to write pretty clearly. their writing is often recognizable and full of tics which voids a lot of the value. its low aura. but I think it’s mostly wrong when people say model writing lacks analytical or informational value

@Jsevillamol Yeah, it's way over 2x now.

There is no way that SWEs are being sped up 2x, but I can believe there is a very solid productivity increase already in place, and that it will become even larger in the future.

@AndrewYang Psychedelics taught me that

My Mom said “don’t hate other people” to me as a kid and I listened.

@MerlijnTrader AI

BREAKING: 🇺🇸 Warren Buffett the greatest investor in history just repeated himself. 1999: "Euphoria is the enemy" Walked away from the rally. Result: dot-com crash. Down 78%. 2026: "We've never had people in a more gambling mood than now." $400B in cash. Zero purchases. Michael Burry: $1B short on AI. "Feels like 1999." Warren Buffett: $400B in cash. "Worse than 1999." Two of the greatest investors alive. Same year. Same warning. Same answer. Cash. Not stocks. Are you listening?

@protosphinx I see you are unfamiliar with AI

The IPO system is broken. You don’t go public to raise growth capital anymore. You do it to dump equity on retail. NVIDIA went public at ~$600mn in 1999. Microsoft at 780mn. Oracle at ~270mn. Intel at 58mn! Today you won’t IPO even at $60 billion. At a $600 million valuation if you hit the jackpot and become a $60 billion company in public markets that’s a 100x return. Investing $10k as retail makes you a millionaire. This is wealth creation. At $60 billion there’s very little chance you get to $600 billion let alone 100x from there. You will certainly lose money as retail. Mathematically if you only allow $100 billion IPOs you’re basically capping upside at maybe 10-20x. There’s no more 100x or 1000x potential. And even 10x is extremely unlikely. Airbnb, Uber DoorDash are maybe 10x outcomes at best for retail investors. These are supposed be success stories of last decade. Apple Microsoft Tesla Google were all 100x opportunities for retail. So you’ve basically cut the average middle class investor out of the upside while private markets and insiders keep most of it for themselves.

@VictorTaelin Tell it to write in the style of your favorite author. Works great

is there hope that GPT models will ever talk normally? it feels like a non-English speaking coworker. or more like I need an autistic-to-English dictionary to navigate it. it almost doesn't matter how smart the model is if the communication is that dysfunctional. really sad

It's not HYPE. Cmon what else you need?

thats crazy

I need you to understand how ridiculous the trajectory is that AI is on

Can somebody find the sigmoid researcher and ask them to update the best fit?

@sama The ability to write a very high quality novel

what would you most like to see improve in our next model?

@__tinygrad__ @ChaseBrowe32432

@ChaseBrowe32432 people see 3 datapoints and fit an exponential. it's an s-curve. every technology is an s-curve. we ran ahead of moore's law mostly due to dtypes.

Mythos lands slightly above the trendline for the AI 2027 scenario

@__tinygrad__ What's the paradox where even highly intelligent people say really dumb things?

Actually it's even worse than the Moore's law about cost cause things aren't getting cheaper! A lot of the AI gains in the last 10 years come from increased spending and smaller dtypes, both of which are nearing an end and we'll be back to the normal Moore's trajectory.

Can we get rid of all the AI hype? Normally I find hype just distasteful, but this is distasteful, filling people with dread, and making hardware expensive. The third one is really annoying. It's just Moore's law and a bubble, there's no singularity.

@__tinygrad__ @ChaseBrowe32432 Whether it's an exponential or an S curve doesn't matter at all when the vertical part of that line is so fucking massive that everything that happens after that vertical is completely alien and incomprehensible to minds that existed before that vertical.

@JoshPurtell Weird take

@BrianCoryDobbs @Mars_Revealer Once again I am reminded why its so important to keep the word retard in the vernacular.

Comment from YouTube... "There are only two explanations possible: 1 - this is not a hexagon. 2 - there was a very old civilization of creatures living on Mars at the time of the Dinosaurs." Discovered by @Mars_Revealer. Full video⬇️

@janleike What's it about?

Some personal news: I am starting a new research project at Anthropic. Very excited about this! Many things are needed to make AGI go well, and alignment is only one of them. More on this soon…

@WarMonitor3 Where's AlexNet, Attention Is All You Need, AlphaGo, and o1?

The Top 10 most impactful moments in history: 1-Julius Caesar Assassination 2-Battle of the Marathon(The West defeats Persia) 3-Fall of Rome ends 1000 year roman order 4-Black Death Arrives in Europe kills half of Europe 5-Columbus reaches the Americas 6-Gutternberg prints the Bible 7-Assassination of Francis Ferdinand starts WW1 8-Dropping of the atomic bomb new nuclear order 9-Fall of the Berlin Wall 10-Hitler invading Poland starting WW2