The North Pole used to be basically unreachable — thick ice for miles in every direction. Now research ships can sail right through waters that used to be frozen solid, and scientists are drilling into ancient seafloor sediment to understand what the Arctic looked like millions of years ago. Meanwhile, AI is being used to analyze all that data much faster than any team of researchers could do by hand. Think of it like having a really fast reading assistant who can skim through thousands of pages of climate history and flag the important patterns while the scientists focus on the actual thinking.
The humanoid robot piece is equally interesting. Companies building robots that walk and move like humans need enormous amounts of data to train them — basically showing the robot thousands of examples of how to pick something up, open a door, or navigate around furniture. Right now a lot of that data comes from humans wearing motion-capture suits and doing tasks repeatedly. It’s slow and expensive, kind of like how you’d teach a toddler to tie shoes by doing it over and over in front of them. The race is on to find smarter ways to generate this training data faster, because whoever solves that problem will have a big head start in building useful robots.
So what does any of this mean for your wallet?
If you run a small business that involves physical, repetitive tasks — packing, sorting, light assembly — keep an eye on what companies like Figure and 1X are doing with humanoid robots. You’re probably 5-7 years away from renting one affordably, but staying informed now means you won’t be caught flat-footed. Subscribe to a free newsletter like this one that tracks the space without the noise.
Climate data analysis tools are quietly getting very good and very cheap. If you’re in agriculture, real estate, or insurance, AI tools that help you assess climate risk for specific locations are already available. Something like ClimateAI or even asking ChatGPT to help you interpret a NOAA report can help you make smarter decisions about where to buy property or how to plan a growing season.
If you’re looking for freelance work, “AI training data collection” is a real and growing job category right now. Companies pay real people to perform tasks, record movements, or label information so their AI models can learn. Platforms like Scale AI and Appen list these gigs regularly and you can do many from home.
The boring-sounding data problems in AI are actually where the real opportunities are hiding.
Join the conversation
Be respectful. Offensive language is automatically blocked.