Ask HN: Likelihood that legislation protects certain industries from AI?

2 points by ngngngng ↗ HN
One of the seemingly more realistic situations I have playing out in my head right now is that as AI starts taking jobs (Artists, Programmers) certain industries manage to successfully lobby the government to make AI illegal to use in their field. Doctors and Lawyers seem like a good bet for jobs that politicians will arbitrarily keep safe.

4 comments

[ 4.6 ms ] story [ 18.8 ms ] thread
Politicians are the first to go.

(1) ChatGPT's most superhuman skill is bullshitting and those who bullshit will be replaced first. (2) Human Feedback Reinforcement Learning is like the ballot box but better. That kind of model can really learn how to tell people what they want to hear at a level far beyond any mortal.

It's a funny bit, but following the Looney Tunes logic lawyers would be next... and I'd rather avoid AI legal council.

Also, the idea of AI democracy escapes me. Why do people think that the sum of all written text will somehow amount to utopia? It's like a mirror image of the people who thought that taking the accountability out of money would fix it's exploitation problem (cough cryptocurrency cough).

I reckon a 0% chance, at least in the US. There's pretty much no way to effectively regulate this, and the opportunity cost of stopping the innovation is probably too great. There's a liability problem, but it dovetails with a money trail so it will get solved.
Roles like lawyers and doctors are already protected from tech. If you made a machine that could administer eye exams without an optometrist (something that has been doable for at least a decade) it would be illegal. Only an optometrist can prescribe prescriptions for eyewear.

Even if AI is helpful in lots of areas a ton of jobs will still require a license for certain jobs.