Comment by reedf1 | Hacker News Reader

Comment by reedf1 | Hacker News Reader

> To me, overusing AI, destroying ecosystems, covering up fuck-ups, and hating minorities are all “bad” for the same reason, which I can mostly sum up as a belief that traumatizing others is “bad”. You cannot prove that AI overuse is “bad” to a person who doesn’t think in this framework, like a nihilist that treats others’ lives like a nuisance.

But you haven't defined the framework. You know a bunch of people for which AI is bad for a bunch of handwavey reasons - not from any underlying philosophical axioms. You are doing what you are accusing others of. In my ethical framework the other stated things can be shown to be bad, it is not as clear for AI.

If you want to take a principled approach you need to define _why_ AI is bad. There have been cultures and religions across time that have done this for other emerging technologies, luddites, the amish, etc. They have good ethical arguments for this - and it's possible they are right.

It's not hard to formulate why "AI" is bad, at least in its current form. It destroys the education system, is dangerous for environment, things like deepfakes drive us further towards post-truth, it decreases product quality, AI is replacing artists and similar professions rather than technical ones without creating new jobs in the same area, it increases inequality, and so on.

Of course, none of these are caused by the technology itself, but rather by people who drive this cultural shift. The framework difference comes from people believing in short-term gains (like revenue, abusing the novelty factor, etc.) vs those trying to reasonably minimize harm.

> The framework difference comes from people believing in short-term gains (like revenue, abusing the novelty factor, etc.) vs those trying to reasonably minimize harm.

OK - so your framework is "harm minimization". This is kind of a negative utilitarian philosophy. Not everyone thinks this way and you cannot really expect them to either. But an argument _for_ AI from a negative utilitarian PoV is also easy to construct. What if AI accelerates the discovery of anti-cancer treatments or revolutionizes green tech. What if AI can act as a smart resource allocator and enables small hi-tech sustainable communes. These are not things you can easily prove AI won't enable even within your framework.

Reply on news.ycombinator.com