Thank you and goodbye @vainglory. The last 4 years have been the time of my life. I travelled around the world, met great people and competed alongside amazing teammates. Thanks to everyone who was part of my journey!
I'm joining forces with @mikeknoop to start Ndea (@ndeainc), a new AI lab.
Our focus: deep learning-guided program synthesis. We're betting on a different path to build AI capable of true invention, adaptation, and innovation.
AGI is the most important technology in the history of the world. That's why I'm going all in on a new adventure with @fchollet. We’re creating a new intelligence science lab -- Ndea.
@ndeainc's ultimate goal is to compress science timelines. First by understanding intelligence, then measuring it, building it, and finally deploying it for scientific advancement.
Our first focus is on deep learning-guided program synthesis to create AGI that can invent, adapt, and innovate.
My view is that in 5-10 years we’ll regard deep learning and program synthesis as equally important to scaling up efficient, bottleneck-free AGI.
The “transformer” for program synthesis has not been created. The "Keras" for program synthesis does not exist.
If this is correct, 2025 looks like 2013. The world just got its first taste of program synthesis when o3 beat ARC-AGI-1, similar to deep learning's moment when AlexNet beat ImageNet.
I mean it when I say “adventure”. Humanity will use AGI to solve incredibly important problems.
But even more exciting is the chance to metaphorically time-travel into the future: learn, invent, and discover things that would not organically happen for decades or even centuries.
Ndea is my focus. So it's time for me to step away from day-to-day at Zapier. I remain a board member, evangelist, and most importantly great friends with @bryanhelmig and @wadefoster.
Zapier’s success is why I can make this bet. I’m extremely grateful to all our customers, employees, and supporters. Zapier has a big mission ahead — bringing AI automation to everyone.
I believe that anyone with new ideas and ability to accelerate AGI should be working on it. The world deserves it.
Chollet and I have tried to support this goal indirectly. We’ve both invested in AGI startups, created @arcprize, written papers, and given talks to inspire open progress. We’ll continue to do all these things.
Now, we're going to work on it directly.
We’re assembling the world’s top program synthesis team. @ndeainc is a globally remote company. Consider joining us!
SFI researcher Alessandro Palmarini focuses on tackling the Abstraction and Reasoning Corpus (ARC) challenge, with the goal of creating AI systems capable of abstract reasoning far more efficiently than existing methods. Inspired by François Chollet’s work on measuring intelligence, his earlier work has built upon the DreamCoder program synthesis system, introducing a novel approach called "dream decompiling" to improve library learning in inductive program synthesis. Listen to Palmirini’s interview on @MLStreetTalk highlighting the delicate balance between computational efficiency and data efficiency in AI learning processes.
#AI #MachineLearning #AbstractReasoning
https://t.co/ThpM4l565a
For those asking if I'll be coming back to play WildRift—I won't be. I stopped gaming altogether at the end of 2019, having felt fully fulfilled with my Esport endeavours. I switched to studying AI and I'm loving it! There are so many exciting possibilities in this field! 🤯
I've never worked so hard in my life and I improved so much as a player these last 10 weeks, both mentally and skill-wise. It wasn't enough to qualify for worlds, but I'm not giving up on @FortniteGame. I truly believe I can become one of the best and can't wait for more events.