Good morning 👋 Today’s reminder to follow your dreams comes from a guy who skipped buying a house to launch a VR-controlled humanoid robot fight club. In a hidden San Francisco warehouse, people use VR headsets and combat controllers to make the robots punch, kick, and sword fight… only sometimes the bots get carried away and destroy their gym. Still beats a mortgage.
⛑️ Major breakthrough in microrobots could revolutionize healthcare. Scientists at Penn State have shown how tiny “talking” robots can self organize using sound waves, like swarms of bees or birds. These simple shape-shifting bots adapt instantly, detect threats, and even heal themselves if damaged. Someday they might act like microscopic first responders – fixing your body from the inside or cleaning up environmental disasters.
🐳 Open source is the fastest way to win the global AI race. China gets it. They’re already spreading their open AI models across Asia, and adoption is booming. Now Silicon Valley and Washington see the threat, and are pushing American-made open models to catch up. Both sides know once a model gets adopted, it’s game over. Switching later almost never happens. This isn’t about who builds the best tech; it’s about who hands out the free samples first.
🫧 Everyone said AGI was just around the corner. But what if this is as good as AI gets? GPT-5 just dropped, and instead of a massive leap forward it delivered modest, incremental improvements. Now even insiders are questioning whether endlessly scaling up models is the right strategy, or if the path to AGI needs a complete rethink. The real question is can we keep the bubble afloat long enough to find out?
🤦♂️ Is AI making doctors worse at their jobs? A new study in The Lancet found doctors became worse at spotting precancerous growths after just a few months using AI. Researchers call it the “Google Maps effect,” where relying too heavily on AI causes doctors to lose their edge and forget crucial skills. Warning: Common side effects of medical AI include forgetting medical school, random acts of malpractice, excessive WebMD searches, and asking patients, “So… what do you think we should do?”
Your AI compliance team
Vanta AI Agent handles compliance tasks automatically. It doesn’t need you to set rules. This agent thinks ahead, learns your system, and spots issues you might miss.
Why does it matter? Startups waste weeks each year on manual compliance checks. Vanta AI Agent catches problems before auditors do, giving your team time back to focus on real work.
Use it to simplify audits and get compliant.
🪼 Nuclear power is back, with billions pouring into new plants all over the US. But before celebrating this clean energy renaissance, we might want to check with our oceanic overlords first. Massive jellyfish swarms have recently shut down nuclear plants in France, Japan, Scotland, Israel, and Sweden. Decades of protests couldn’t stop the power, but warm ocean Jell-O just might.
🚀 WTF is Q2T3? Silicon Valley insiders just dropped a freakish new hypergrowth metric for AI startups. It means hitting $100M ARR in as little as four years. Some are even doing it in just 1.5 years. VCs aren’t betting on unicorns anymore, they’re breeding mutants.
fun stats
amount of AI unicorns 🦄 worth $2.7 trillion collectively.
AI is minting billion-dollar startups faster than the dot-com era. 100 unicorns were born in the last 2 years – including OpenAI, Anthropic, and Scale AI.
[Fortune]
Want massive geek cred? Build this radio-controlled Lego-powered submarine with automatic depth control