AI future is clearly local, and my recent pitch has been "infinite tokens." Because that's what my M1 MBP can do; and that's what my RTX3090 can do. I don't need to pay hundreds of dollars a month and no one else does either.
In the 80s we thought that the future of computing was clearly local, home computers, PCs, Macs, the office server (Novell, then Windows NT with disk shares) etc. Add 40 years and we are back to a centralized infrastructure with the modern equivalent of smart terminals.
The AI future will be clearly... what it will be. Probably bouncing back and forth from local to centralized. However, if there are money to be made by selling things that people run locally, it seems that centralizing creates more power and hence more money.
the real money is in the coding surrounding models to make them efficient at specialized tasks. Casual users want general purpose models, and AI chat apps will stay for them. Most programs can benefit from a specialized AI that can be local, and #programs >> #users.
The AI future will be clearly... what it will be. Probably bouncing back and forth from local to centralized. However, if there are money to be made by selling things that people run locally, it seems that centralizing creates more power and hence more money.