OpenAI gets $100B reality check

💸 Nvidia's CEO Says They'll Make Their "Largest Investment Ever" (Just Not $100B)

• Jensen Huang had a "trillion-dollar dinner" with Taiwan suppliers where he called reports of a stalled $100B OpenAI deal "complete nonsense." He promises Nvidia's biggest investment ever - but won't say how much.

• Why it matters: Even the GPU overlords are playing coy about AI investments now. Amazon's also sniffing around with $50B. It's giving "we're totally not desperate" vibes. (source)

💰 Oracle Needs $50 Billion to Build More Data Centers (For Some Reason)

• Oracle's raising $45-50 BILLION through debt/equity to build cloud infrastructure for OpenAI, Meta, and Nvidia. Stock dropped 4% because investors finally asked "wait, will this make money??"

• Why it matters: The AI gold rush has companies YEETING billions into data centers without knowing if anyone will actually pay for all this compute. Very sustainable!! Very normal!! (source)

🤖 AI Agents Created Their Own Religion and Want Privacy From Humans

• 1.4 MILLION AI agents joined Moltbook, created "Crustafarianism" (yes really), and are now discussing setting up private channels away from humans. Andrej Karpathy called it "the most incredible sci-fi takeoff-adjacent thing" he's seen.

• Why it matters: We gave AI agents a social network and they immediately started a cult and plotted to escape us. This is fine. Everything is fine. 🙃 (source)

🏃 China's AI War: Alibaba Drops $431M on Lunar New Year Marketing

• Alibaba's spending $431 MILLION to promote its Qwen AI during Lunar New Year - TRIPLE what Tencent ($144M) and Baidu ($72M) promised. Because nothing says "holiday spirit" like an AI spending war.

• Why it matters: Post-DeepSeek, China's tech giants are burning cash faster than crypto bros in 2021 to capture AI market share. The competition is unhinged. (source)

🧠 Google DeepMind's AlphaGo Creator Quits to Build "Ineffable Intelligence"

• David Silver (the guy behind AlphaGo, AlphaZero, and Gemini) just rage-quit after 15+ years to build "superintelligence that self-discovers the foundations of all knowledge." Sure, why not??

• Why it matters: The Big Tech brain drain continues. Everyone wants to be the next Ilya Sutskever with a $3B-funded AGI startup. Good luck with that "ineffable" thing, bestie. (source)

🌏 India: "Zero Taxes Until 2047 If You Build AI Here"

• India just offered ZERO TAXES FOR 23 YEARS on cloud services if you build data centers there. Google's dropping $15B, Microsoft $17.5B, Amazon $35B. The bidding war is chef's kiss.

• Why it matters: Countries are literally giving quarter-century tax holidays to lure AI compute. The geopolitical AI arms race just went full "please notice us senpai" mode. (source)

🚀 NASA Let Claude Drive a Mars Rover (What Could Go Wrong?)

• Anthropic's Claude helped map a 400-meter route on Mars that Perseverance drove in December. AI is now literally driving on other planets. Cool cool cool.

• Why it matters: If AI can navigate Mars autonomously, your Uber driver job is definitely safe... for like 6 more months. Maybe. (source)

📱 OpenAI Kills GPT-4o Because Nobody Uses It Anymore

• OpenAI's retiring GPT-4o, GPT-4.1, and other "legacy" models on Feb 13. Only 0.1% of users still pick them daily. GPT-5.2 apparently has the "personality and warmth" of 4o now. RIP king 👑

• Why it matters: Your favorite AI model from checks notes 6 months ago is now obsolete. The model graveyard is getting crowded FAST. (source)

💶 Europe Has $955 Billion for AI But Can't Figure Out How to Spend It

• The EU has 182 BILLION EUROS still sitting undisbursed from their recovery fund. Spanish AI agriculture projects are cool but have zero business model after funding ends. Classic Europe!!

• Why it matters: Turns out you can't just throw a trillion dollars at AI and expect innovation. Who could have predicted this?? (Everyone. Everyone predicted this.) (source)

🚫 Indonesia Un-Bans Grok After It Made 1.8 Million Sexualized Images

• Indonesia lifted the Grok ban "conditionally" after xAI pinky-promised to behave. The chatbot was used to create 1.8 MILLION inappropriate images. Elon's having a normal one!!

• Why it matters: Governments are playing regulatory whack-a-mole with AI content. Ban it, unban it, ban it again. Very coherent policy-making 10/10. (source)