From OpenAI to Open-WHY?

Welcome back apprentices! 👋

AI’s Top Talent Is Pulling a Houdini — And Their Former Bosses Are Sweating!

Picture this: You’re at a high-stakes poker table, and suddenly, half the best players just up and leave — only to set up their own casino across the street. That’s exactly what’s happening in the AI world right now. 

Top researchers and executives from OpenAI, DeepMind, and other tech behemoths are vanishing faster than your phone battery on a bad day. But they’re not retiring to grow organic avocados — they’re building their own AI empires.

Why the mass exodus? Are they tired of corporate politics? Chasing something bigger? Or do they just want to prove they can do AI better than their old bosses? 

One thing’s for sure: the AI industry is shifting gears fast.

So grab your popcorn (or maybe your morning coffee), because this isn’t just another tech shake-up — this is the AI revolution in full swing.

In today's email

• From “Revolutionary” to “Returned”

• AI Tries Freelancing

• NBA Unveils AI-Powered Robots

• A Recycling Revolution?

• AI Can Now Tell If Your Dog Loves You

• The Great AI Exodus

• More Tools & Updates

_{Read Time: 6 minutes}

Quick News 

🙃 The Humane AI Pin was hyped as the “screen-free future”, making it onto Time’s Best Inventions list and winning RedDot’s Best Innovative Product award. Fast forward to 2025, and HP just bought Humane for $116M — without the AI Pin. Turns out, the future wasn’t here after all — more Pins were returned than purchased, and reviewers called it so bad that its purpose was impossible to understand.

💡 OpenAI’s new SWE-Lancer benchmark put AI to work on 1,400 real Upwork coding tasks with $1M in payouts at stake. The best model, Claude 3.5 Sonnet, completed almost half the tasks, raking in a theoretical $400K — but no AI walked away a six-figure freelancer just yet.

🏀 The NBA’s All-Star Tech Summit unveiled AI-powered robots like A.B.E., a rebounding bot used by Stephen Curry, and M.I.M.I.C., which runs plays like a never-tired teammate. Other bots motivate players, inflate basketballs, and organize gear — basically, the world's most high-tech waterboys.

♻️ Sorting plastic waste is a costly, labor-intensive nightmare, but researchers at The National Institute of Standards and Technology (NIST) may have a fix. Their new AI-powered laser sorting system uses near-visible infrared (NIR) light to rapidly scan and identify plastic types, making recycling faster, cheaper, and more efficient.

🐶 Scientists are training AI to read animal facial expressions, meaning your dog’s side-eye and your cat’s perpetual judgment may finally have scientific explanations. From smart farms scanning pigs for stress to an upcoming app that detects cat pain, AI is proving better than humans at understanding animals — because let’s be honest, we’ve all misread a tail wag before.

Together with The Rundown

Learn AI in 5 minutes a day

This is the easiest way for a busy person wanting to learn AI in as little time as possible:

1. Sign up for The Rundown AI newsletter

2. They send you 5-minute email updates on the latest AI news and how to use it

3. You learn how to become 2x more productive by leveraging AI

Sign up to start learning.

The Great AI Exodus 
Why Ex-OpenAI Leaders Are Going Rogue and Building Their Own Empires 

If AI had a Wild West, we’d be in it right now. The sheriffs of OpenAI, Google DeepMind, and other tech giants are losing their deputies faster than you can say "alignment problem." And where are these top AI minds going? They’re not retiring to the mountains. They’re launching startups — big, bold, and sometimes directly competing with their former employers.

The latest defector? Mira Murati, OpenAI’s former CTO, who just unveiled her new venture, Thinking Machines Lab. And she’s not alone. Anthropic, Safe Superintelligence Inc., Adept AI, and Inflection AI — all founded by former OpenAI executives — are already making waves.

But why are these brilliant engineers and researchers jumping ship? And could one of these new startups dethrone OpenAI as the AI kingpin?

What’s Cooking?

Mira Murati, one of the master architects behind ChatGPT, has officially announced her new AI venture, Thinking Machines Lab. No more whispers or vague job postings — this thing is real, and it’s aiming high. While the details are still emerging, one thing is clear: Murati wants to change the way AI is built, understood, and used.

Instead of just improving chatbots or squeezing out marginal improvements in text generation, Thinking Machines Lab is focusing on making AI more customizable, widely understood, and deeply integrated into scientific and programming fields. 

Murati isn’t just tweaking AI — she’s attempting to rewire how it interacts with human intelligence itself.

What We Know About Thinking Machines Lab

🧪 A Science-First AI Approach 
This isn’t about making AI that just automates customer service calls. Murati’s company is diving into frontier AI research — focusing on science, programming, and real human-AI collaboration. 

The goal? To create AI that doesn’t just spit out responses but actively assists in research and problem-solving across various disciplines.

🧠 Hiring an Army of Ex-OpenAI Researchers
If AI had an Avengers team, Murati is assembling it. Alongside OpenAI veterans like John Schulman and Barret Zoph, she’s also bringing in top minds from DeepMind, Character AI, and Mistral. Essentially, she’s building an AI dream team, and they’re here to disrupt the game.

🔥 A Not-So-Subtle Challenge to OpenAI
Murati isn’t outright declaring war on OpenAI, but let’s be honest — when a former CTO suddenly launches a competing AI lab and fills it with ex-OpenAI talent, the message is loud and clear: “We think we can do this better.” The startup’s focus on openness and human-AI collaboration could position it as a stark contrast to OpenAI’s increasingly corporate, closed-door approach.

💡 A Possible Open-Source AI Future?
One of the biggest criticisms of today’s leading AI companies? Their secretive, locked-down research. Thinking Machines Lab is flipping the script, committing to regularly publishing technical papers, code, datasets, and model specs. This means the broader AI community — startups, academics, and independent researchers — could benefit from their breakthroughs.

 🤑 A Billion-Dollar Valuation Incoming?
If there’s one thing investors love, it’s AI startups with ex-OpenAI leadership. Given the massive VC appetite for AI (with SSI already in talks to raise over $1 billion), Thinking Machines Lab is almost guaranteed to hit unicorn status in record time. And they haven’t even shipped a product yet.

Why Are AI Employees Leaving Big Tech? 

AI research is in its rebellious teenage phase. The top minds in the field are slamming their doors, yelling “you don’t understand me!” at their corporate bosses, and launching their own ventures. 

Why? A mix of frustration, ambition, and the undeniable lure of huge wads of cash.

Key Reasons for the AI Exodus:

🏛️ Corporate Red Tape is Annoying
Big AI companies have enough red tape to gift-wrap the moon. Researchers want to experiment, build, and break things, but instead, they’re drowning in meetings and compliance checklists.

🛡️ AI Safety vs. AI Profit Wars 
Some ex-OpenAI researchers, like Ilya Sutskever, left to build AI that won’t accidentally kickstart Judgment Day. Others are just tired of watching their bosses chase profits instead of prioritizing the safety of, you know, the entire human race

💸 Investors Are Throwing Cash at AI Startups
With $50 billion+ invested in AI startups in 2024 alone, why take orders from a boss when you can get venture funding and be your own boss? AI researchers are taking the hint and cashing in.

🚀 Startups Move Faster Than Big Tech
Large corporations move at the speed of a grandma crossing the street. AI startups? Jetpack mode engaged. Researchers who crave agility and the freedom to innovate are fleeing big tech for nimbler, riskier, but more rewarding startups.

The New AI Mafia

Murati isn’t the only one. OpenAI has basically turned into a talent factory for future AI disruptors. Here are some of the biggest names making waves:

• Anthropic (Founded by ex-OpenAI’s Dario and Daniela Amodei) — These guys raised billions from Amazon and Google, and they’re putting AI safety at the forefront (translation: they want AI to be helpful, not homicidal).

• Safe Superintelligence Inc. (SSI) (Founded by Ilya Sutskever)—This one’s literally just about building “safe superintelligence”, so we don’t get wiped out by our own creations.

• Adept AI (Founded by OpenAI engineers) — Instead of just building chatbots, Adept is creating AI that can actually perform tasks for you. Backed by Nvidia.

• Inflection AI (Founded by OpenAI & DeepMind vets) — They built Pi, an AI chatbot designed to be more… emotionally intelligent. (Because apparently, that’s what AI needed?)

Could One of These Startups Become the Next AI Superpower? 

Tech history shows underdogs often win. Google took down Yahoo, Tesla outmaneuvered legacy automakers, and Netflix obliterated Blockbuster (RIP).

Now? One of these AI startups might just outthink, outmaneuver, and outfund OpenAI.

With billions in backing, the best talent in AI, and a whole lot of “let’s prove them wrong” energy, these startups are setting the stage for an epic battle. Will Thinking Machines Lab emerge as the AI kingpin? Or will Anthropic, Adept, or SSI steal the throne?

One thing’s for sure — the AI wars are just getting started. 

Tools & Updates 

⚛️ Microsoft just dropped Majorana 1, a quantum computing chip powered by mysterious topoconductor material (yes, that’s a real word). The goal? To cram a million qubits onto a single chip, finally making quantum computing actually useful — because apparently, all the world’s computers combined aren’t cutting it. But here’s the twist: scientists are still debating whether Microsoft’s magic particles even exist.

🌊 Meta just announced Project Waterworth, a 50,000 km undersea mega-cable because apparently, the internet needs its own deep-sea highway. This multi-billion dollar oceanic flex will supercharge AI, boost global connectivity, and make sure businesses in the U.S., India, Brazil, and South Africa can stream, compute, and meme faster than ever. 

💼 Fiverr just dropped Fiverr Go, an AI toolkit that lets freelancers train AI on their own work — because who doesn’t want a robot apprentice? For $25/month, you can sell AI-made versions of your work, and for $29/month, an AI assistant will handle client emails so you don’t have to pretend to be polite. Oh, and top freelancers? They’re getting company shares, because why not add “shareholder” to your gig worker resume?

🧪 Google just unleashed an AI-powered co-scientist (built on Gemini 2.0) that generates and tests scientific hypotheses faster than your over-caffeinated research intern. In trials at Stanford and Imperial College, it identified new drug applications and gene transfer mechanisms in just days—because who has time for slow science anymore?

🔬 Arc Institute and NVIDIA just launched Evo 2, an AI model so powerful it ingested the entire genetic code of life — literally 9 trillion DNA building blocks from 128,000 species. It can predict cancer-causing mutations with 90% accuracy, design synthetic genomes, and probably tell you if your morning smoothie has good genes.

📱 Apple has introduced the iPhone 16e, a cheaper, AI-powered sibling to the iPhone 16, packing the A18 chip, a 48MP 2-in-1 camera, and battery life that lasts longer than your willpower at a dessert buffet. With AI perks like Genmoji, Clean Up in Photos, and Siri that actually listens, it’s designed to make premium features more affordable.

📰 The New York Times is bringing AI into the newsroom, letting tools like GitHub Copilot, Google’s Vertex AI, and OpenAI’s API handle SEO, summaries, and research — basically, the digital equivalent of an unpaid intern. They’ve even built Echo, an AI summarization tool, but AI is strictly banned from writing articles or making images (because, let’s be honest, no one wants a robot covering politics just yet).

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