The North Pole’s Ice-Free Past and the Humanoid Data Gold Rush

Two stories caught my eye today, and they couldn’t be more different in scale but equally fascinating.

First up, the Arctic. Last year, a research vessel sailed straight to the North Pole through open water and thin ice. That’s not supposed to be easy. It used to mean grinding through meters-thick ice for weeks. The fact that it was a cakewalk is a stark reminder of how fast the Arctic is changing.

Now scientists are drilling deep into the seabed to answer a big question: was the Arctic Ocean ever completely ice-free? And if so, what does that tell us about where we’re headed? It’s not just academic curiosity. The answers could help predict how quickly the northernmost waters might lose their ice cover again. Tim Kalvelage has the full story in the latest print issue, which is all about nature. Worth a read if you care about what’s happening to the planet.

On the other end of the spectrum, there’s a weird new trend in AI: companies paying people to film themselves doing boring everyday tasks. I got invited to an app that would pay me to microwave food or put things in bowls. Another site wanted me to remotely control a robotic arm. What’s going on?

It’s the humanoid data gold rush. Robotics companies are desperate for real-world movement data to train their machines. As James O’Donnell explains, your mundane actions are becoming valuable training fodder. It’s one of the 10 Things That Matter in AI Right Now, and honestly, it’s both clever and a little creepy.

Meanwhile, the AI spending race is getting absurd. Google, Microsoft, Amazon, and Meta collectively set new records, spending 71% more than the same quarter last year. The NYT has the numbers. Microsoft, Google, and Amazon are seeing payoffs, but Meta’s shares slid after investors got spooked by their plans. Makes you wonder about that AI bubble everyone keeps talking about. MIT Tech Review has a good piece on that.

The White House is pushing back against Anthropic’s plan to expand access to their Mythos model, citing cyber risks. Bloomberg has the details. They’re also worried about losing government compute access. Meanwhile, Anthropic is reportedly seeking funding at a valuation over $900 billion. That’s a lot of zeros for a company that’s still figuring out safety.

Elon Musk testified that OpenAI’s leaders “looted the nonprofit” and called himself “a fool” for trusting them. Gizmodo has that. But let’s not forget Musk had his own concerns about OpenAI. The drama never ends.

There’s more in the must-reads, but those are the highlights. The Arctic drilling and the humanoid data push are the stories I’ll be following closely.