I suspect that people would prefer answers by AI *even if they were actually written by a human".
Social animals, above a certain intelligence, are wary of deception by members of the same species. Computer programs, websites, and now AI, work round that by appearing to be neutral and impersonal. Which means that paradoxically we seem to end up trusting distant organisations more than those that actually have a greater degree of shared interest.
Imagine proposing, in a vacuum, a future with opacity and autonomy stripped from all internet search and research.
It's inconceivable that people accept giving up their own agency because the technology behind corporations now controlling the agency is impressive. "Open source" is not going to keep this from hurting the world just like the ability to cook at home didn't save the health care system from McDonald's.
Yeah, I realize that my comment could be taken as implying that people's "revealed preference" is to lose their autonomy to AI. No, that would be a bad thing. What I meant was really that people fail to recognise the potential for adversarial behaviour in seemingly impersonal automatons, which is a problem that predates AI. But people can learn.
I suspect that people would prefer answers by AI *even if they were actually written by a human".
Social animals, above a certain intelligence, are wary of deception by members of the same species. Computer programs, websites, and now AI, work round that by appearing to be neutral and impersonal. Which means that paradoxically we seem to end up trusting distant organisations more than those that actually have a greater degree of shared interest.
Imagine proposing, in a vacuum, a future with opacity and autonomy stripped from all internet search and research.
It's inconceivable that people accept giving up their own agency because the technology behind corporations now controlling the agency is impressive. "Open source" is not going to keep this from hurting the world just like the ability to cook at home didn't save the health care system from McDonald's.
Yeah, I realize that my comment could be taken as implying that people's "revealed preference" is to lose their autonomy to AI. No, that would be a bad thing. What I meant was really that people fail to recognise the potential for adversarial behaviour in seemingly impersonal automatons, which is a problem that predates AI. But people can learn.
Management replaces developers with AI: this is good, AI is magic and developers are monkeys.
AI replaces management: this is bad, management is magic and AI is monkeys.