My 6200 Likes & Retweets

tpot is just aspiring counter elite twitter https://t.co/cDAR1bOKvJ

This could be an incredible revolution in Cosmology. The Dark Energy model of the universe, which won a Nobel Prize in 2011, may be completely wrong. The accelerating expansion instead is simply because time runs faster in the voids between galaxies. Let me explain: https://t.co/kAiE9aWYFV

i want to grow my ideas like a garden. a conceptual prototype inspired by this thread of tweets. https://t.co/sOYKmdimbN https://t.co/AtU5C6ZcIk

Here’s how I would prepare for AGI: Work is going to become a lot less important. If you are the sort of person who ties your self-worth to what you do, I would start decoupling that. We might live for a very long time. You are going to want to be as healthy as possible. Get…

Today OpenAI announced o3, its next-gen reasoning model. We've worked with OpenAI to test it on ARC-AGI, and we believe it represents a significant breakthrough in getting AI to adapt to novel tasks. It scores 75.7% on the semi-private eval in low-compute mode (for $20 per task… https://t.co/ESQ9CNVCEA

New verified ARC-AGI-Pub SoTA! @OpenAI o3 has scored a breakthrough 75.7% on the ARC-AGI Semi-Private Evaluation. And a high-compute o3 configuration (not eligible for ARC-AGI-Pub) scored 87.5% on the Semi-Private Eval. 1/4 https://t.co/uQA47JWkl6

Day 12: Early evals for OpenAI o3 (yes, we skipped a number) https://t.co/iWXg9IGuZM

2025 will be the year of agentic systems The pieces are falling into place: computer use, MCP, improved tool use. It's time to start thinking about building these systems. At Anthropic, we're seeing a few best practices emerge - we wrote a blog post with our findings: https://t.co/rlJhVoJWgI

Nature just published a piece on why probability doesn't exist. Lee Cronin said many times chemical reactions are not real. Your genetics professor probably confessed to you before that there are no genes. Biologists still disagree on whether cell types are even real. And we… https://t.co/7JEKDucaxY

In tldraw computer, you create workflows of connected components that generate and propagate data. Components can take any othe…

Everything you love about generative models — now powered by real physics! Announcing the Genesis project — after a 24-mon…

Community notes are too slow to matter much. As Twain didn’t say, “a lie can travel halfway around the world while the truth is still putting on its shoes” https://t.co/1aGhngGm71

venn diagrams are so back https://t.co/CmleVXIFGk

Everything you love about generative models — now powered by real physics! Announcing the Genesis project — after a 24-month large-scale research collaboration involving over 20 research labs — a generative physics engine able to generate 4D dynamical worlds powered by a physics… https://t.co/4HRaTp5Gbs

We just launched tldraw computer https://t.co/KRtNAWDtJT

software engineers have an unfair advantage in the AI age not because of their coding knowledge but because they know you need to thoroughly describe and break down a problem before trying to solve it most people are monkeys that throw random darts at a problem until one hits https://t.co/wItSSmP8du https://t.co/ROkcq7qysn

Preprint out today that tests o1-preview's medical reasoning experiments against a baseline of 100s of clinicians. In this case the title says it all: Superhuman performance of a large language model on the reasoning tasks of a physician Link: https://t.co/QB0mawHJNE A 🧵⬇️

I tried so hard to be useful to the world and make real impact. I was masking my true self and pretending to be all these things I am not — a mechanical engineer, a roboticist, a control theorist, a computer vision researcher. I got quite far in this performance and the world…

If you're feeling unemployable... If you're feeling like a misfit... Just employ yourself and do your own things. Create your own job, start your own company. Believe in yourself, you can do the things! https://t.co/hrI3lRCQk8

Visualization and sonification 🎹 of the @github commit history of the @scikit_learn repository. https://t.co/bdvE8wkB7U

Today we announced Veo 2: our state-of-the-art video generation model which produces realistic, high-quality clips from text or image prompts. 🎥 Sign up to try Veo 2 in VideoFX ↓ https://t.co/xNap881O5a https://t.co/4Z9UzErBhX

Lottielab is now open to everyone 🔥 https://t.co/ZhOhulFcmz

I'm not saying this is the TikTok algorithm. I'm just saying this is a 2022 Recommendation System paper from Bytedance - the parent company of TikTok https://t.co/cQga3L7bk0

Only the unimaginative limit themselves to hypotheses when analyzing a new dataset. A thread on exploratory data analysis ⬇️ https://t.co/Ep9LAfH3jq

I’m unemployable because I'm a generalist. I like to do a bit of everything but I don't have depth in the things I do. Things like product, engineering, design, sales, marketing, and customer support. Sadly, there isn't a position called "Generalist" at companies! https://t.co/vahlbSsQbZ

After 2 weeks of working on the Miniperplx roadmap, the search groups feature is finally ready and live! try it here → https://t.co/obprjeZwTY github → https://t.co/lPDX1cohau all features, except writing mode, are powered by @ExaAILabs! also launching it on uneed[.]best… https://t.co/94PE8KMIGG

A new paper from Meta gets rid of tokenization by learning a transformer over raw bytes https://t.co/fpgqZqaVMr

One of the best points that someone made to me once was this: "Humans generalize on far less data than AI currently does. That means there's something our brains are doing algorithmically to do far more with far less data. Until we figure out that paradigm, we are no where near… https://t.co/IoaRRQv28y

Before you can have “fuck you money” , you gotta have “fuck you skills”… …and before that you gotta have “fuck you courage”… …and before that you gotta have “fuck dreams”… …and before all that you gotta be able to look at everything and everyone in the face and say “fuck…

META JUST KILLED TOKENIZATION !!! A few hours ago they released "Byte Latent Transformer". A tokenizer free architecture that dynamically encodes Bytes into Patches and achieves better inference efficiency and robustness! (I was just talking about how we need dynamic… https://t.co/aawPMIRjjG