Contrary conjecture: As with many "productivity enhancements", these AIs[0] will streamline some work to the point of practical elimination, but will in the en

Contrary conjecture: As with many "productivity enhancements", these AIs[0] will streamline some work to the point of practical elimination, but will in the end make more work, by enabling higher-level tasks to be doable instead of just imagined. It seems a lot of the missing "productivity enhancements" from the PC and internet revolutions don't appear because people didn't net get fired, they just started doing different tasks (TBD whether those are actually useful, but evidently most managers think they are). Oblig car analogy - more highways induce demand for travel by making it easier, so the net is that traffic is worse; similarly, this may induce demand for work. [0] these are artificial, but not intelligent; it is predictive models with insane amounts of ingested data. Whether text, code, or images, it speeds production in middle-of-the-road cases, but cannot figure out what to produce. There's no more evidence of reasoning than that an image in a mirror is evidence of skilled kinesiology. With an actual reasoning AGI, the conjecture would be quite different.