AI and open source
i keep being right about private predictions about where AI stuff will be in ~2 years time so here's a public one
the AI bubble pops, many of the big name companies go under, less (but not none, not even close) of them get deployed in commercial products
meanwhile the "open source" models continue incrementally improving. more and more open source software devs start turning the way of Cory Doctorow and champion this as a victory, begin integrating it as an indispensible tool in their lives. their skills atrophy. the models get incrementally better, but not that much better. already large and complex codebases inflate in size by comical amounts.
open source software goes to shit, and the reputation of open source software follows. the people who actually know what they are doing are hired by companies doing whatever the next hype thing is, since at least that way they aren't constantly flooded by poor-quality AI code that looks fine at first glance but has deeper structural problems.
successful open source projects in this future probably look a lot more cathedral than bazaar, with teams of devs who know and trust each other doing everything privately.
the AI bubble pops, many of the big name companies go under, less (but not none, not even close) of them get deployed in commercial products
meanwhile the "open source" models continue incrementally improving. more and more open source software devs start turning the way of Cory Doctorow and champion this as a victory, begin integrating it as an indispensible tool in their lives. their skills atrophy. the models get incrementally better, but not that much better. already large and complex codebases inflate in size by comical amounts.
open source software goes to shit, and the reputation of open source software follows. the people who actually know what they are doing are hired by companies doing whatever the next hype thing is, since at least that way they aren't constantly flooded by poor-quality AI code that looks fine at first glance but has deeper structural problems.
successful open source projects in this future probably look a lot more cathedral than bazaar, with teams of devs who know and trust each other doing everything privately.