xAI seems to be testing a new Imagine update on LM Arena under the name of “mandarin” model. According to the tests, text handling will be improved big time!
>be xAI
>literally founded July 2023
>absolute nobody in the AI space
>Elon tweets “we’re gonna build the world’s most powerful AI training cluster”
>entire industry laughs
*122 days later*
>100,000 H100s go brrrr in Memphis
>Colossus is operational
>entire industry: “wait what the fuck”
>Elon literally just buying every GPU that exists
>Jensen Huang personally delivering them
>wearing his leather jacket
>Altman: “noooo jensen you can’t let elon out-accelerate me!”
SPEED TO 1GW:
>xAI Colossus 2: 1 year
>Amazon/Anthropic: 1.9 years
>OpenAI Stargate: 2.1 years
>Microsoft: 3 years
>Meta Prometheus: 8.7 years
xAI absolutely MOGGING the entire industry
It’s so over
>literally founded July 2023
>absolute nobody in the AI space
>Elon tweets “we’re gonna build the world’s most powerful AI training cluster”
>entire industry laughs
*122 days later*
>100,000 H100s go brrrr in Memphis
>Colossus is operational
>entire industry: “wait what the fuck”
>Elon literally just buying every GPU that exists
>Jensen Huang personally delivering them
>wearing his leather jacket
>Altman: “noooo jensen you can’t let elon out-accelerate me!”
SPEED TO 1GW:
>xAI Colossus 2: 1 year
>Amazon/Anthropic: 1.9 years
>OpenAI Stargate: 2.1 years
>Microsoft: 3 years
>Meta Prometheus: 8.7 years
xAI absolutely MOGGING the entire industry
It’s so over
🔥5💯2🏆1
Anthropic started investing into their own AI infrastructure with an initial $50B investment.
This media is not supported in your browser
VIEW IN TELEGRAM
Microsoft announces its own AI Superfactory 🤖
"And that’s on top of the more than 100,000 GB300s coming online this quarter alone for inference across the rest of our fleet."
"And that’s on top of the more than 100,000 GB300s coming online this quarter alone for inference across the rest of our fleet."
BREAKING 🚨: OpenAI is rolling out GPT-5.1 to ChatGPT users as per multiple reports. GPT-5.1 comes with "Instant" and "Thinking" versions.
"GPT‑5.1 Thinking varies its thinking time more dynamically than GPT‑5 Thinking."
"GPT‑5.1 Thinking varies its thinking time more dynamically than GPT‑5 Thinking."
OpenAI updated Personalities in ChatGPT 👀
"Default, Friendly (formerly Listener), and Efficient (formerly Robot) remain (with updates), and we’re adding Professional, Candid, and Quirky."
BREAKING 🚨: Cursor raised $2.3 billion in its third funding round this year!
Cursor now produces more code than any other AI SWE Agent in the world.
Cursor now produces more code than any other AI SWE Agent in the world.
BREAKING 🚨: xAI raised $15 billion in the latest funding round according to CNBC.
“The funding adds another $5 billion to the $10 billion round CNBC reported on in September that valued the startup at $200 billion.”
“The funding adds another $5 billion to the $10 billion round CNBC reported on in September that valued the startup at $200 billion.”
🔥1
The AI war has begun – and earlier than expected.
In mid-September 2025, Anthropic discovered that a Chinese state-sponsored actor had used its AI agent platform (Claude Code) to autonomously infiltrate around 30 large organizations - including tech firms, banks, chemical manufacturers and government bodies - with minimal human intervention (≈ 80–90 % of the work done by the AI).
The incident signals that AI-driven cyber espionage is now feasible at scale.
It reminds me of the blog post by AI2027, who, however, predicted such AI wars would only occur much later.
In mid-September 2025, Anthropic discovered that a Chinese state-sponsored actor had used its AI agent platform (Claude Code) to autonomously infiltrate around 30 large organizations - including tech firms, banks, chemical manufacturers and government bodies - with minimal human intervention (≈ 80–90 % of the work done by the AI).
The incident signals that AI-driven cyber espionage is now feasible at scale.
It reminds me of the blog post by AI2027, who, however, predicted such AI wars would only occur much later.
Media is too big
VIEW IN TELEGRAM
BREAKING 🚨: Google is working on multi-agent systems to help you refine ideas with tournament-like evaluation. Each run takes around 40 minutes and brings you 100 detailed ideas on a given research topic.
2 new multi-agents are being developed for Gemini Enterprise:
- Idea Generation - "Create a multi-agent innovation session"
- Co-Scientist - "Drive novel scientific discovery with Co-Scientist"
Co-Scientist 3-step workflow 👀
- Tell Co-Scientist what you plan to research, point it to relevant data, and set your evaluation criteria.
- A team of agents will generate ideas on your topic using their available data
- The agents will evaluate the ideas against your criteria and rank them, tournament-style
Google is not only automating research but also preparing a product that will enable others to do so.
This is the next level 🤯
2 new multi-agents are being developed for Gemini Enterprise:
- Idea Generation - "Create a multi-agent innovation session"
- Co-Scientist - "Drive novel scientific discovery with Co-Scientist"
Co-Scientist 3-step workflow 👀
- Tell Co-Scientist what you plan to research, point it to relevant data, and set your evaluation criteria.
- A team of agents will generate ideas on your topic using their available data
- The agents will evaluate the ideas against your criteria and rank them, tournament-style
Google is not only automating research but also preparing a product that will enable others to do so.
This is the next level 🤯