For starters, I'd love to see someone with hiring or legislative power tackle the subject of displacement without being dismissive.
For example, anyone who dedicated a bit of brainpower could understand that prompt engineering was bullshit from the beginning. Yet today we're still seeing it being rehashed as an alternative to knowledge work in the form of harnessing, guardrailing or context engineering. This is scaleable neither size-wise (how many different wrapper-builders do you really need?) nor time-wise (how long until model providers vertically integrate that, too?).
The reactions that we see today - the calls to ban AI or implement UBI, or even the race to escape the "permanent underclass" - are some symptoms of this lack of direction, IMO.
Personally, I'd like to see:
- Automation tax in line with displacement, which is then directly funneled to services and fund pools that mitigate the effects of displacement. If you want to reduce the winner-takes-all outcome, you should provide the capital to many more to own the automation tech, robots, whatever.
- Firing moratoria as an industrial policy tool. IIRC, the Chinese implemented something of this sort: you can automate, but you have to retain a certain amount of jobs. Obviously, it's a transitional solution.
- Aggressive expansion of software- and model-auditing & engineering jobs and education. If this is the direction, more people should be able to understand the systems in detail, not fewer.