1 - Within 2 years most jobs will become "managing a bunch of robots". Within 3 years that part won't need mentioning. The job is whatever the actual domain/product/service is, again.
2 - Cheap AI will reduce the need for capital at least as much as it reduces the need for labour (per unit of production). So self-employment and much smaller businesses will become viable in many fields.
3 - There will be more than one brand of self-driving car available for purchase by consumers in the next 12 months, which will cost less than four times a regular car.
4 - Self-driving cars will gradually become more like lounge-rooms, then tiny apartments, then slightly less-tiny apartments. Service stations and carparks will evolve to support this.
5 - Most governments will have a sovereign (hosted/owned in their country), quite decent ai chatbot within 12 months and it will hook up to government data/sites/services and be free for all citizens
6 - Within 12 months the question of "Was any AI at all used in this?" will become very rare for writing, code, visual art, movies, music, etc, and will return to "Is it any good?"
7 - Higher education will revert to questioning people in conversation to test what they know. They'll also test people in realtime on how well they use, review, criticise, and judge the output of AI
8 - General purpose humanoid robots (doing 3 or more different, useful tasks fairly well) are less than 12 months from market and most people who can afford them will want more than one
9 - A free llm that runs on the average (not prosumer/gamer) laptop and beats current (Sept 2026) frontier-models, will exist within 12 months. Probably by better separating a core reasoning engine from an (accompanying) knowledge-base
[link] [comments]