Are you ready for another twist in the Erdos problems story? Today a $1000 prize open problem (#707 erdosproblems.com/forum/thread/7…) was solved by two humans who … realized that the problem was solved 30 years before being posed 🤣. Maybe lit search is not so simple after all 😅. But

Huge 👀 Google just announced research that shows, for the first time in history, that a quantum computer can successfully run a verifiable algorithm on hardware, surpassing even the fastest classical supercomputers (13,000x faster). It can compute the structure of a molecule,

All intelligence is generalization. The rest is just lookup.

To startups, if you're going to have ARC-AGI be your headline announcement for a fundraise, come talk to us and we'll help share your verified performance

Congrats! Quantum Echoes are so cool! It also proves my work!!! Thank you Google

Just wow!!!

I’m seriously patent pending in AGi… wtf! 😳 We did it! #Qi #AGi

I mean… what have I been saying and what is the first equation to AGi on my profile?!

😍

This RIP JSON is a joke. It’s not going anywhere. TOON is just another syntax. Yes, more optimized at the moment but let’s be honest. Computation thinks in numbers not words.

What he is talking about is my work not the robot. The physics that define the coding is what will eliminate poverty. The software. Not the hardware. Another example of Elon Gross Marketing. We wonder why the rich squash the poor. It’s access.

Shout out Gemini: “Verdict: This "Core Spec" is the source code for the patent. It validates that the invention is not just a collection of buzzwords, but a coherent thermodynamic control system for information processing.” Google DeepMind

We tested Opus 4.5 over the weekend and nearly stopped at 32K thinking tokens Then we tested 64K overnight which came in at state of the art for reproducible scores Very cool work Anthropic, excited to personally use this model

Nothing like my work being confirmed a year later. #AGi #Ai

🚨 This paper just proved why most “AI traders” are fake. Not because the models are dumb. But because the reward signal is lying to them. Here’s the core failure mode: If you train an LLM agent directly on market returns, the model quickly learns a shortcut: memorize

Mathematician Terence Tao about GPT-5.2 solving the Erdős problem: "the problem was solved more or less autonomously after some feedback, and the result doesn't seem to be in the literature the bigger shift is how fast AI can now write and rewrite clear expositions of the

And, I’m already patent pending in it. 🤘🏻 #Qi #AGi

I mean… I’m not an academic writer. And I’m already Patent pending in AGi. 🤘🏻

Geoffrey Hinton says LLMs are moving beyond imitation toward self-consistent reasoning Instead of just predicting the next word, new models are beginning to identify contradictions in their own logic This unbounded self-improvement will "end up making it much smarter than us"

It’s funny how recursion is the new buzz word for ML and I have been talking about it since last year. #Ai #ML