the Microdose

AI Face Off

+ agents evolve, demo bots, and Amazon gets caught
Adam Wildheart

AI Scarlett Johannson deepfake video/The Microdose

Subscribe to The Microdose
Subscribe -- post
Cheri Wildheart
Adam Wildheart

Good morning 👋 to everyone in tech whose dating pool suddenly upgraded from “nerdy” to “absurdly hot and curious about confidential projects.” Passion is exciting, but ask yourself why your date’s pillow talk involves trade secrets. Some claim Russia and China are deploying operatives to seduce tech pros for info using “sex warfare.” Insiders see a surge in suspiciously attractive strangers sliding into LinkedIn DMs, sometimes even marrying their victims. But before you interrogate your partner, consider dropping fake intel and enjoying the ride.

🙅‍♂️ The next big AI battle is over who owns your face. AI-generated deepfakes can replicate anyone’s face and voice, and current laws aren’t ready. Celebrities have already lawyered up, federal legislation is brewing, but deciding who controls your digital likeness isn’t easy. Critics worry new protections might overreach, forcing websites to remove legit content like parody and satire. If lawmakers mess this up, your digital twin could end up with more rights than you.

🎓 For AI to reach superintelligence, the student must become the teacher. Building bigger models won’t make AI smarter, no matter how many billions companies like OpenAI and Meta throw at them. Thinking Machines Lab argues that true superintelligence requires AI that learns and remembers, not tackling isolated tasks and instantly forgetting them. Real breakthroughs will only happen when AI actively learns from experience and continuously adapts the way people do. Without that, we’re not building intelligence, just expensive autocomplete machines.

✨ Most people miss the real play behind AI browsers. They’re evolving into agents. Every click, scroll, and purchase teaches AI how to act more like you. Meanwhile, the browsers subtly shape your habits, guiding you to depend on them. They’re not passive windows to the web anymore. They’re pressing buttons, filling forms, and quietly retraining you. Pretty soon, you’ll wonder who trained who.

🤜🤛 Together with

Create AI agents easily

Lindy AI agents help you automate the repetitive, boring stuff. Describe what you want, connect your apps, and within minutes you’ll have custom AI agents working for you. 

Why you’ll love it: It’s like hiring intelligent 24/7 teams who never miss a deadline. Whether you’re building solo or scaling up, Lindy has everything you need to automate anything from sales to support, easily.

Use Lindy to grow faster without headcount.

🤖 What if you never had to clean your house or cook dinner again? Domestic humanoid robots like Figure 03 put on impressive demos, but Meta’s chief AI scientist says we’re still years away from that reality. Today’s bots can handle repetitive tasks like folding towels or dancing, but throw real-life chaos at them and they’re completely lost. Robots lack the ability to continuously learn from sensory inputs like touch, motion, and video, so they can’t form intuitive “world models.” Until they figure out how to learn, managing life remains your responsibility.

🚁 The biggest bottleneck in drone warfare isn’t AI, it’s hardware. Ukraine’s drone program makes that painfully clear. Despite deploying thousands of drones with increasingly smart targeting systems, full autonomy is still out of reach. The problem isn’t software. Small drones simply don’t have enough computing muscle to fly solo, especially when remote connections fail in combat. For now, Ukraine keeps people firmly in control, proving that no matter how smart AI gets, physics still has the final say.

💧 Wait, Amazon lied about its datacenter water use? Nobody saw that coming. To polish its sustainability image, Amazon publicly claimed its datacenters use just 7.7 billion gallons of water per year. But leaked documents reveal the actual number almost 14x higher at 105B gallons. Scientists say Amazon’s splashy “Water Positive” campaign is nothing more than greenwashing wrapped in creative accounting. But hey, can we really call it lying if regulators won’t call them on it?

fun stats

9.3%

of recently published news was written by AI ✍️ Smaller outlets in the south and mid-Atlantic use AI the most (some above 20%), and it’s rarely disclosed. 

Only 1.7% of articles in major papers like the WSJ, NYT, and Washington Post had AI generated content, with 6.4x more of it sneaking into opinion pieces written by guest contributors.

🔮 “If definitive proof emerges that we’re living in a simulation, the correct response to that news will be, So what. A life lived in a simulation is a real life.” – Sea of Tranquility by Emily Mandel

subscribe today

Get an edge with The Microdose

Skip the prompts, get the signal. We’ll send you short daily updates about the real AI + future tech you need to know. Fast, smarter, mildly addictive – and free. 

Subscribe -- page