Unfortunately, You Need to Know What the Jevons Paradox is
One thing I think a lot AI boosters get wrong is that they think AI will be good at creating new, quality information while, at the moment, the only thing it has shown utility at is organizing existing information. Even the weights themselves are a kind of distillation of existing information.
Much of science...perhaps even the great majority of it, is not organizing existing information, it is acquiring new information. AI might make that more efficient, but it does not do it.
This is one of the things that writing this video made me think. I look forward to other people's thoughts in the
