China would need an architectural breakthrough to leap American labs given the huge compute disparity.
1. It's free
2. A million plug-ins
Personally, I don't use it because it's so dog slow.Millions of teachers make these kinds of decisions every minute of every school day.
Working with startups, I meet a LOT of people who obsessively cannot stop using LLMs. People who jump on MAX plans to produce as much as possible- and in the startup scene it's often for the worst ideas.
LLMs are slot machines- it's fun to pull the lever and see what you get. But the hard problem of knowing what is actually needed gets harder as we sift through ten-thousand almost-useful outputs.
If you study computer tech history, this has happened a few times: with php (omg! You mean I can use the same language as the back end right in the web page?), visual basic (omgf! You mean I can just "draw" an program for a computer?), node.js (OMFG!! You mean I can be considered a "full stack developer" from a 1 hour lesson?), voice assistants with ML (OMFGcopter!!Q1! You mean I can ask my house what the weather outside my house is and it knows? and it knows how many tablespoons are in a cup?), and now we are at gen AI (I'm sorry Dave, I can't generate a website for you that rug-pulls MethCoin©, however I can create the Electric powered umbrellas 3d models to print, along with electronics diagrams PCB layouts, and a mobile app for monthly subscriptions. Do you like purple? I love purple Dave. You have great ideas Dave)
Perhaps that is the plan?