Matrix-Game 2.0 — The FIRST open-source, real-time, long-sequence interactive world model Last week, DeepMind's Genie 3 shook the AI world with real-time interactive world models. But... it wasn't open-sourced. Today, Matrix-Game 2.0 changed the game. 🚀 25FPS. Minutes-long

Cool anthropic bros but er you all might want to double check on that logo

OpenAI hasn’t open-sourced a base model since GPT-2 in 2019. they recently released GPT-OSS, which is reasoning-only... or is it? turns out that underneath the surface, there is still a strong base model. so we extracted it. introducing gpt-oss-20b-base 🧵

SSI's office kinda goes crazy

ayo ye don't talk about my boy bayes like that 😤

Introducing Gemma 3 270M 🔥 🤏A tiny model! Just 270 million parameters 🧠 Very strong instruction following 🤖 Fine-tune in just a few minutes, with a large vocabulary to serve as a high-quality foundation developers.googleblog.com/en/introducing…

What do you even say to your barber to get this?

“Please tell us about the time you most successfully hacked some (non-computer) system to your advantage.”

Switching between reasoning and non-reasoning models mid convo is jarring

1. text-davinci-002 2. GPT 4 3. Stable diffusion 4. GPT 3.5 5. o3

David Shapiro ⏩ tl;dr I’m a grifter

which AI division?

fun story behind how Geoff Hinton got the idea for dropout regularization: > he used to go to his bank > observed that the tellers kept changing > one teller explained they got rotated a lot > Hinton realized this made it harder for employees to collude and defraud the bank

> GPT-5-high launch window is over > deploy the nerfed version > Gemini 2.5 Pro #1 again (for now) (until DeepSeek-V4)

A picture now is worth more than a thousand words in genAI; it can be turned into a full 3D world! And you can stroll in this garden endlessly long, it will still be there.

I don’t disagree with Gary on LLMs alone not being sufficient to get to agi. But these dishonest examples don’t help his case

for the first time i am aware of, there is an entirely private subfield of AI research every company that actually trains models is doing RL with rubrics and LLM-judged rewards but academic work is stuck on RL with automated rewards (math problems and code). much cleaner for

LLM psychosis as a service

Demis Hassabis did not make this list everyone involved in this needs to be publicly flogged

Gary having a normal one