🔥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

Crazy that early stage AI is so competitive that PearVC offering $250k uncapped SAFE + $600k cloud credits to AI PhDs/Profs Marc Andreessen 🇺🇸 and his barbell VC theory, where the seed/series A/B funds get wiped out and only megafunds or pre-seed funds survive seems more and more credible

By my Oxford friends Josh Lavorini and @tobyloader ! Implemented via web ZK proofs of deposits via apps ppl use like Revolut, instead of some weird crypto wallets The onramp/offramp & wallet embedded into a single unified interface, as a bet against normie stablecoin adoption

Becoming an RL diehard in the past year and thinking about RL for most of my waking hours inadvertently taught me an important lesson about how to live my own life. One of the big concepts in RL is that you always want to be “on-policy”: instead of mimicking other people’s

Second richest man in the world says he will "concentrate resources" into Ellison Institute at Oxford. Yeah, all you Americans are right, Europe is shit. Do not bother coming. Nothing to see here. 😂

Current AI bottlenecks are not capabilities; it's adaptive learning. One of the most interesting toy examples that exemplifies where AI are not superhuman is exploitative poker; Noam Brown already solved the MinMax GTO (game theory optimal) poker, where you optimise around not

KPOP is a new DL optimizer designed for large scale distributed training on Apple Silicon. KPOP uses a lot more memory but is more efficient per FLOP than AdamW, so it's a better fit for hardware with a high memory:flops ratio. Some hardware numbers: H100: 80GB, 1000TFLOPS

Jamie Dimon is the Jensen Huang of Finance Silverback hair, intense and calm as hell simultaneously, running an overnight success playbook for decades

The new Slack API rate limits are absolutely insane: For accessing messages, you can now only make 1 request per minute, with a maximum of 15 messages 🤯 And yes this applies to ALL apps, not just public ones. We can't build internal tools on our OWN data. Marc Benioff this is

How is no one talking about the fact that THE Veo3 was easily crushed by bytedance's new video model??????!??? Is this some weird CCP psyops

Couldn't believe it when I saw this Google is losing search market share so badly they have to start putting ads in London underground First time in my life I saw a Google ad

When flocks of Technology Brothers in SF flood Anthropic with a bunch of requests through Claude Code ,and you can't even ask basic questions anymore ☠️☠️☠️

Do feel like European founder victim mentality of there being no VCs who take a risk on you is overplayed Especially for seed ecosystem I think Europe is fine, sure its not as low friction as SF but if you get rekt by friction you probably shouldn't do startups anyway

LLMs are a "Chinese person locked in a cave" - "Chinese Room" argument of genuine understanding vs rule-centric response - Plato's cave argument of the sim2real gap of ingesting the entire internet but never having touched grass and interacted in the real world