Comment by w10-1

Solo and small-group computer projects used to be in a sweet spot of opportunity: feasible yet interesting to others. But opportunity-space is limited and fills.

To me a more important question is: where can people 10-30 years old bootstrap themselves on interesting and useful problems? They have intelligence and energy but not resources or connections (mostly), and all that potential human capital will all be wasted if they don't have any tractable and fruitful domains. (We don't have to worry about those with resources, connections, or luck, but they're a small minority already tethered to value flows, in little danger of being under-developed.)

My concern in particular is that tech companies spent big on building free languages and tools (yes, you used to have to pay for compilers, IDE's, databases...) in order to reduce input costs of them and their customers. If AI already minimizes labor costs (both the work and the discovery and training of residual human talent), there's no reason to subsidize that self-training, and likely fewer portable skills (across more isolated tech stacks), further locking employees in, reducing cross-pollination (formerly within the valley).

Individual opportunity is shrinking. Young people feel it. Old people feel it too, even if they have bagfuls of tech stocks.