Google DeepMind formed a strike team to improve its coding models, with Sergey Brin directly involved.
It’s surprising to me that Google has the world’s largest internal codebase (>2B LOC), yet lags behind Anthropic and OpenAI in coding + agents.
“Google’s AI writes 50% of code, trailing Anthropic’s near 100%.” Google engineers are largely limited to internal models and tools (like Gemini CLI and Antigravity), so this makes sense.
I think Sergey in founder mode can fix it again this time.
> be me
> build "AGI" benchmark
> actually version 3 already
> we don't talk about 1 and 2
> (they saturated in a year)
> invent new scoring method
> if human scores above AI, use squared efficiency
> example: human took 10 steps to solve level
> AI took 100 steps to solve a level
> then AI's score is (10/100)
> correction, add square penalty to be extra cheeky
> (10/100)^2
> AI scores plummet
> smug_face.jpg
> observe that AI is scoring higher than human on some levels
> "damn"
> "CLAMP THE SCORES!!!"
> new metric now takes minimum of 1 and AI's score
> AI literally NOT ALLOWED to score higher than human
> meanwhile human get squared rewards
> doubly_smug_face.jpg
> AI thinks at 200 tokens/s
> human thinks at 5 tokens/s
> cries about it
> "UNFAIR ADVANTAGE"
> add restriction that models can't use more than 5X the steps that a human used
> DeepThink, Pro and xhigh models exist that think longer to score better
> excluded them
> DECLARE AI HIT A WALL
Introducing TurboQuant: Our new compression algorithm that reduces LLM key-value cache memory by at least 6x and delivers up to 8x speedup, all with zero accuracy loss, redefining AI efficiency. Read the blog to learn how it achieves these results: https://t.co/CDSQ8HpZoc
1/4 LLMs solve research grade math problems but struggle with basic calculations. We bridge this gap by turning them to computers.
We built a computer INSIDE a transformer that can run programs for millions of steps in seconds solving even the hardest Sudokus with 100% accuracy
The Pentagon and at least one Gulf government are in talks with Ukraine to buy Ukrainian drone interceptors -FT
The countries are looking to Ukraine's low-cost options for shooting down Iranian attack drones.
Ohhh nooo not my private IP how dare someone use that to train an AI model, only Anthropic has the right to use everyone elses IP nooooo, this cannot stand!
It’s 2028. You’re on the run from the authorities. Full blown AI psychosis consumes you. You don’t give a fuck, you finally did it. You stole a Claude.
24 dedicated people.
$30M spent on development.
Extreme specialization, speed, and power efficiency.
Today we launch Taalas’ first product. Check it out:
Details: https://t.co/88CA0XAL71
Demo chatbot: https://t.co/ec4ladcKnw
API: https://t.co/M3EkaxEqPj