Form: Short Note

(Thoughts via Sr Fellow Francis Zhou)
RE: Economic consequences of the AI technology:
– Humans have a tendency to overestimate the effect of a new technology in the short term, and underestimate its effect in the long term. The “crisis” is more likely to playout in slow motion for the next 20 years, til 2048, instead of 2028.
– In any human organization of sufficient size, it isn’t technology advancement that determines the pace of progress, but one hiring and firing/retirement at a time, especially at the upper echelons. This takes far longer than typically expected, which explains “organizational inertia”.
– +1 on the difference between vibe coding to create a working prototype, vs operating a SaaS service with scale, reliability, and compliance. For most enterprises, software is a utility. You subscribe to the utility, you don’t aim to own the infrastructure that produces the utility, unless you have to.
– Moving bits have became cheaper and cheaper first following Moore’s law, then Metacalf’s law, to the point now that some unnamed AI law will make it even cheaper. Moving atoms doesn’t benefit from these laws. So if job security is a concern, go into a career that involves moving atoms. It doesn’t have to be “blue collar” per se, building robots for example, consists of moving atoms. At least until general purpose robotics pair with AGI becomes an affordable reality (I think we are at least a few more decades away from that reality).
– Capital doesn’t get destroyed outside of war, it just flows to another sector/asset.