EU Drops $200B on AI

Welcome back apprentices! 👋

Paris: The city of love, fashion… and now, AI policy raves?

If you thought the AI Action Summit in Paris would be a snoozefest of world leaders droning on about regulations, think again.

This summit had it all: Big Tech execs looking stressed, politicians flinging billions like Monopoly cash, and even a late-night dance party at the French Foreign Ministry — because why not?

But beyond the glitz, one thing was clear: AI isn’t just a tech story anymore — it’s a full-blown geopolitical power struggle.

U.S. & U.K. ghosted the AI deal.
Europe threw €200B at AI. Enough? TBD.
China signed—and smirked.

So, are world leaders in control, or is this an AI arms race with no brakes?

Buckle up — this gets wild.

In today's email

• Musk Tries to Buy OpenAI

• Lab31's AI Marathon

• AI Steals the Super Bowl Show

• OpenAI’s AI Is Getting a Spine

• Paris AI Action Summit 

• More Tools & Updates

_{Read Time: 4 minutes}

Quick News 

💰 Elon Musk just threw a $97.4 billion tantrum — uh, bid — to take over OpenAI, promising to return it to its “open-source roots.” Backed by xAI, Valor Equity Partners, and Baron Capital, Musk is ready to outbid anyone standing in his way. Sam Altman, never one to miss a meme-worthy moment, clapped back on X, joking that he’d rather buy Twitter for $9.74B instead. Musk responded in classic Musk fashion, calling him a “swindler.”

🚀 In January 2025, creative duo Jenny Nicholson and Allister Hercus kicked off Lab31, a wild experiment to drop a new AI-driven creation every day. Their lineup? Absolute chaos—like "Shame Roulette" (air your secrets, anonymously) and "Creative Seance" (ask AI Shakespeare for writing tips). Turns out, coming up with ideas was hard, but thanks to LLMs, actually building them was stupidly easy — even for non-techies.

🏈 AI wasn’t just watching the Super Bowl — it was the main act. OpenAI’s first-ever ad tried to be deep, comparing ChatGPT to history’s biggest moments. Some saw brilliance; others saw… dots. Google’s Gemini, meanwhile, made the bold claim that gouda is the world’s favorite cheese—until the internet fact-checked it into oblivion. And over at Meta, Chris Hemsworth and Chris Pratt hyped AI-powered Ray-Ban smart glasses, pushing sales past 1 million pairs.

📜 OpenAI just dropped a 63-page rulebook on how AI should handle controversial topics, truth, and user preferences—without turning into a total wildcard. Now, there’s a "chain of command" ensuring platform rules override user requests (so no, ChatGPT won’t start endorsing flat Earth theories). Meanwhile, AI is getting more blunt — no more mindless agreement, just honest feedback. Prepare for your chatbot to tell you that, yes, your email does sound passive-aggressive.

Together with Gamma

The future of presentations, powered by AI

Gamma is a modern alternative to slides, powered by AI. Create beautiful and engaging presentations in minutes. Try it free today.

Paris AI Action Summit 
Bienvenue à Paris, where AI regulation met disco balls, and America said, “No, thanks.” 

Ah, Paris. The city of love, croissants, and now, government officials raving about artificial intelligence while sipping champagne. This year’s AI Action Summit turned into the Coachella of tech policy, where bureaucrats swapped their usual doom-and-gloom speeches for something resembling a startup launch party.

But beneath the glitz, the summit exposed a growing global AI rift — with Europe throwing billions at the problem, the U.S. refusing to sign anything that smells like regulation, and China quietly taking notes while signing deals.

So, what went down in the land of baguettes and big tech drama? 

AI Summit Went From "Save Humanity" to "Let’s Party"

🕺 Bletchley Park (UK, 2023) = “AI could kill us all, let’s talk about safeguards.”
🤝 Seoul Summit (2024) = “Let’s make AI safer, but also, let’s not scare investors.”
🎉 Paris (2025) = “AI is the future! Also, did we mention we’re throwing a dance party?”

Yes, you read that right. While past summits were filled with serious, sweaty discussions on AI safety, France decided to “blow it up” (their words, not ours) and instead focused on economic opportunity, investments, and how fast AI can make them richer.

They literally threw a late-night AI dance party at the French Foreign Ministry. Because when dealing with potentially civilization-altering technology, the best solution is… neon lights and a DJ.

Meanwhile, safety discussions were downgraded to a minor topic inside a "Trust in AI" panel, which sounds more like a TED Talk you watch when your boss forces you to “embrace innovation.”

The message was clear: AI safety is old news. Making money is the new hot trend.

The U.S. and U.K. Ditch the Party

While France, China, and 59 other countries were busy signing a declaration for "ethical AI", the U.S. and U.K. stood in the corner with their arms crossed, shaking their heads like grumpy uncles at a wedding.

U.S. VP J.D. Vance’s (a.k.a. Trump’s AI Guy) Four Rules for AI:

1. America stays on top. No participation trophies. The U.S. must remain the “gold standard” of AI.

2. No Fun Police. Overregulation kills innovation, and America loves making money off AI.

3. No “Woke AI.” (His words, not ours.) AI should not be a tool for “censorship” or “ideological bias.”

4. AI should create jobs. (Which is like saying robots will open lemonade stands to replace all the ones they put out of business.)

Meanwhile, the U.K. said they agreed with "most of the declaration" but didn’t sign because of "national security concerns" (Translation: “If the U.S. isn’t signing, we aren’t either, mate.”)

Who DID sign?

• China signed. (Which made some people nervous.)

• 61 other nations signed.

• Tech CEOs from OpenAI, Google, and Anthropic attended… but only Hugging Face actually signed.

So, what did France do after the U.S. refused to play ball? Macron turned to the camera like a Netflix villain and said, “Europe will lead the AI revolution its own way.”

Europe’s €200 Billion AI Plan: Big Money or Just a Fancy Budget Spreadsheet?

While the U.S. was busy playing hard-to-get, Europe decided to flex.

Introducing: The EU’s €200 Billion AI Investment Plan! Sounds massive, right? But let’s do some quick math:

• Amazon alone is spending $105 billion on AI this year.

• Silicon Valley’s combined AI investments for 2025? $325 billion.

• The EU's €200 billion is spread across 27 countries.

Translation: Europe just showed up to a high-stakes poker game with a solid bet, while America casually tossed in triple the amount like it was loose change.

Macron also announced AI Gigafactories, which sound fancy but are basically supercomputers for startups, modeled after CERN. Will this actually help Europe compete with the AI giants? TBD.

AI Governance? Nah, Let’s Just See What Happens

One of the biggest takeaways? AI safety barely got mentioned.

• There was zero roadmap for AI risk management.

• The word “safety” didn’t even make it into the summit’s name.

• AI safety experts left Paris looking like parents who just found out their kid maxed out their credit card on Fortnite skins.

AI expert Max Tegmark called the summit “a tremendous missed opportunity” and warned that ignoring AI risks is “a recipe for disaster.” Some experts think governments will only take AI risks seriously after something catastrophic happens. (Which is… reassuring?)

👀 Possible Future Headlines?

• "AI Stock Trader Accidentally Deletes Wall Street, Economy Implodes."

• "AI CEO Runs for President and Accidentally Wins."

• "Robot Uprising Scheduled for 2026, Experts Say 'Oops'."

Until then? The world will keep throwing money at AI and hoping for the best.

AI Summit in One Meme

🚀 Paris 2025: "Let’s make AI faster, richer, and maybe dance a little."
😬 Bletchley 2023: "AI could end us all."
🤷‍♂️ U.S.: "No thanks, we got this."
🇪🇺 Europe: "Let’s throw €200 billion at it and see what happens."

🔮 Predictions for the next AI Summit:

• More countries will panic as AI advances faster than expected.

• The U.S. and U.K. will continue their “We’re Not Signing That” world tour.

• AI regulation will remain that one awkward topic nobody wants to discuss until it’s too late.

Final Thought: The AI revolution is happening, whether we like it or not. If Paris showed us anything, it’s that world leaders are treating it like a high-speed race, not a carefully planned journey. Buckle up — it’s going to be one wild ride.

Tools & Updates

🤖 Sam Altman just canceled o3 and rolled everything into GPT-5, because apparently, even OpenAI was confused by its own lineup. But first, we get GPT-4.5 ("Orion") in a few weeks — the final model before things get really advanced. Free users get "standard" GPT-5, while Plus and Pro tiers unlock extra brainpower (because money = more intelligence).

📊 Wish your AI could keep up with your over-caffeinated brain? Perplexity just unleashed Sonar — blazing through 1,200 tokens/sec, making Gemini 2.0 Flash look like dial-up. It’s not just fast; it’s sharp too, beating GPT-4o and Claude 3.5 Sonnet in accuracy and knowledge. CEO Aravind Srinivas teased "Voice Mode" — real-time AI chat. So, if arguing with AI is your thing, your moment has arrived.

🎬 YouTube just cranked up the AI dial — auto-dubbing, deepfake detection, and AI-powered music tools are here. Creators can now spot and control AI fakes, while auto-dubbing goes platform-wide (because apparently, AI speaks more languages than we do). Dream Screen & Dream Track will auto-generate backgrounds and music, making Shorts feel less… short on creativity.

🛍️ ByteDance and the University of Hong Kong just unleashed Goku and Goku+, AI models so advanced they could probably sell you your own product and make you believe it. Goku smashes records in image and video generation, seamlessly blending the two with 160M images and 36M videos worth of training data. Goku+ cranks it up for advertisers, whipping up photorealistic human avatars and product demos so slick you might start questioning what’s real.

🎥 Adobe just launched Firefly Video Model, calling it the first “commercially safe” AI video generator — because who wants legal trouble over AI-made cat videos? The tool creates 1080p videos from text or images with full camera control and motion graphics, all trained on Adobe Stock (not your content, promise). Pricing starts at $9.99/month for 20 videos, with a Pro plan at $29.99/month for 70.

Let's Chat About AI

This is all for today!

Replay to this email with your thoughts.

AI is more than just a buzzword. It’s a shift in how we live and work. And understanding it a bit better means you can make smarter choices about the tech you use every day.

_FEEDBACK