Twitter Beğeni Hilesi

What's really going on in machine learning? Just finished a deep dive using (new) minimal models. Seems like ML is basically about fitting together lumps of computational irreducibility ... with important potential implications for science of ML, and future tech...

If you are looking to conduct research full-time on the foundations of AI, and • you have read the RL textbook and done the exercises, • you agree with the Alberta Plan for AI Research, • you already have a PhD, • you are open to spending some time in Edmonton, then the

Non recurrent neural networks unroll the periodic activation patterns of space fixated brains into a layered function application. This leads to a different optimal representation of the computational process, and perhaps to different learning and generalization behavior.

Modeling a lion is easy: just take a model of everything and remove all that is not part of the lion.

Full house at our Xhope salon on positive futures from AI to space with Foresight Institute's Lydia LaRoux, Joscha Bach, Alder, amongst others. Subscribe to our Luma calendar for an overview of all of our in-person events: lu.ma/foresightinsti…

Tomorrow at 1pm, I will give a talk on Machine Consciousness, at Camp 7th Planet (9:45 Intrigue), at Burning Man!

Absolutely incredible.

Everyone is freaking out over this, but there are many logical explanations. Although Starliner is *most* likely haunted, it could also be aliens, space whales, sophon interference, forever lost cosmonauts banging on the hatch begging to be let in, or simply the Babadook.

Marc Andreessen 🇺🇸 I think the concern is that rn most of the world doesn't make room for babies and small children, so mothers become incredibly isolated. And then I'm so tired of my new mom friends being cheated on, new dads complaining abt a stay at home wife when it's literally an

I did some coding this weekend. Trying different ways to work with AI. Verdict: It's awesome for people like me that are rusty but know many of the eng concepts. If you cannot read and fix code yourself you will hit a wall you cannot get through. Haven't had this fun for years.

This weekend, the @xAI team brought our Colossus 100k H100 training cluster online. From start to finish, it was done in 122 days. Colossus is the most powerful AI training system in the world. Moreover, it will double in size to 200k (50k H200s) in a few months. Excellent

upset at the amount of work i'm about to sink into my CHAPPiE Roan costume for halloween but it's all worth it when I think about the 9 favs and 2 retweets it'll get and, of course, the opportunity to spend 3-5 parties explaining the concept of Die Antwoord to people in their 20s

I was asked (via Nathan 🔍) to give a read as to the likely impact of SB 1047 on AI. so for what it's worth: Will SB 1047 cover many models? Something like "Will SB 1047 cover models from more than 10 companies in 2029?" Cover meaning "covered models" as in the bill. I

Excited to announce our Series A! We raised more than $100M to grow Sakana AI into a World Class AI Lab in Japan. We’re going to really push the frontiers of what’s possible with AI. As a founder mode startup, we operate much faster than most frontier AI labs at a global level.

Mountain: identified. Time to climb

Right now 3 out of 4 replies to this tweet are LLM-generated (#1, #3, #4). Long-term this is an existential threat for the web.

What I Got Wrong About "Shock Therapy" After witnessing a historic financial collapse in the former Soviet Union, I thought bad advice from Western economists was a root cause. Now I think failure was the plan open.substack.com/pub/taibbi/p/w…

I get the point & take it as the compliment intended, but will clarify that my vehement long-standing anti-Putin, pro-Ukraine positions are fundamentally pro-Russian! A peaceful, democratic Russia freed of mafia rule, ignorance, and the imperialist virus, able to join the world.

How does TIME put Helen Toner and Scott Wiener on the list but forgets Alexandre de Moraes?

Telegram quietly updates FAQ, removing: "All Telegram chats and group chats are private amongst their participants. We do not process any requests related to them."