🔥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

So cool! I need to get back into robotics, I’ve got to build a bunch of these to work in the lab day and night! Just need the hands ☺️

After seeing this study, Gary Marcus must be having nightmares! 🤣 By the way, Ilya was right about LLMs, he saw this…

While I am very critical of Geoffrey Hinton’s AI doomerism & ideological stance, I nonetheless greatly respect him for his pivotal contributions to advancing AI. Importantly, he also seems like a genuinely good person without a self-serving agenda, unlike some others we know…

This is good, but I would also like to see reviews of the rejected papers if the authors consent. That way, we can use AI to analyze them and assess the fairness of the review process. Additionally, all accepted papers should be made freely available shortly after publication.

AI image of the day. Created with ChatGPT.

Self-driving (AI) cars will eventually save tens of thousands of lives & prevent over two million injuries every year in the US alone! As a secondary benefit, they’ll also greatly reduce insurance costs & prevent massive economic losses. The beneficial impact of AI is incredible!

I believe what I would call “baby AGI” has already been achieved internally by one of the frontier labs, I won’t mention which one I think of, but people can probably guess. I still predict it will likely be released this year, possibly as some sort of autonomous agentic product.

This is a very interesting protein design model, it’s the first time I’m hearing about it. It would be cool to try it out.

We absolutely need something similar here in the US. It seems the UK will lead in AI-based drug discovery with this effort: Announced today, the UK’s ‘OpenBind’ consortium will use breakthrough experimental technology to generate the world’s largest collection of data on how

My question is: why does something like a “McKinsey report” even exist when Deep Research could produce a far better one in an hour for just a few dollars? Think about how much time & money was wasted on obsolete stuff. It’s time for humanity to wake up to what’s happening!

Another remarkable advance from Google AI: they just announced Gemini 2.5 Flash, which scores 11% on humanity’s last exam, at a very low cost! A lite version for just pennies per million tokens. We’re finally witnessing the dawn of “intelligence too cheap to meter”!

I completely agree with this! Even when GPT-3.5 was released, I thought AGI was still a decade away. But with ChatGPT-4, I revised that to 5-6 years. Then with o1-preview I brought it down to 2. After o3, I was convinced we’ll achieve it this year. This wasn’t obvious at all!

A new era in filmmaking has begun!

This AI short film made with Veo 3 is simply incredible! If this is possible now, just imagine what will be possible next year! When OpenAI Sora was released early last year, some of us predicted this, yet many were incredulous that AI could ever produce full-length films.

This is an incredible public dataset, 8.4 million single cells with 20,000 protein perturbations! Truly a game-changer for BioAI! Congrats to Xaira therapeutics and Bo Wang and team!

“AI will discover new science in the next 5-10 years” Sam Altman As a scientist, 100% agree with Sam Altman on this, except I believe AI will make scientific discoveries sooner than five years! Just wanted to go on record for that.☺️

Fun ChatGPT image prompt: “Make an image that reflects my maturity level based on conversations we had” Here is mine, surprisingly AI thinks I am pretty mature 😅 I like the symbolism though😊