3 ms·
The problem with "AI" and endusers is we humans are flawed enough to have a large and only semi-effective science and industry focusing on what amounts to inter
by VLM 14y ago
The problem with "AI" and endusers is we humans are flawed enough to have a large and only semi-effective science and industry focusing on what amounts to interface failures between "intelligences". Abnormal psych, couples counseling, that sort of thing.
True, you do need to worry about the engineering and stats and this might be close to solved for some trivial problems.
Yes you also need to worry about the interactions between AI and "normal" people and this is nowhere near solved even for trivial problems but its been slowly improving for decades.
The biggest problem is debugging interactions between AI and "AB-normal" people. How should the AI react when rubbed up against a OCD person, or a psychopath, or a developmentally disabled enduser?
This I believe to be the fundamental failure mode for AI in enduser products, probably enforced by the greedy legal system. If you ignore the most vulnerable members of the population you knowingly released a product that kills them, thats not going to turn out well. Or you can hyperoptimize it such that your lawnmower is better at dealing with psychopaths than the smartest human, in which case its hyperregulated by the medical system up to unaffordable cost.