Discussion
Disaggregating the marketing term "AI" is good, but so is aggregating the eugenicist motivations for measuring and promoting different technologies in terms of some supposedly measurable quality of intelligence.
Part of the trouble of doing so, though, is avoiding also aggregating artistic and theatrical senses of the term "AI" as seen in, for example, behavioral models used to create the illusion that NPCs in games are real.
The moment you try to measure intelligence as a thing that can be synthesized and manipulated, you have more overlap with euginicists than is probably comfortable. There's ways of navigating that without adopting eugenicist framing, but that's not a project that Silicon Valley capital classes seem interested in.
It's not coincidental that Bostrom, one of the more influential philosophers in the SV / AI space, peddles anti-Black racism about white people being inherently more intelligent.
If you pose an intelligence metric by which some AI product or technique is measured, you'd damn well better be able to explain why it's a category error to apply your metric to humans!
Install bonfire.cafe
Get the full app experience
