three possible intuitions about the near-term impact of LLMs on hiring and job performance:
1) the standard "technology doesn't eliminate jobs, it creates more of them, and we're all more productive!" line
2) a small number of people with the right skillset (non-ironically, good "prompt engineers") take over everything; everyone else loses out
3) LLMs are a great equalizer: since all the hard work is done by AI, almost any human with a randomly good idea can be successful
mechanistically, these are all mutually compatible, though they of course push towards very different outcomes.
i think to varying degrees, and in varying domains, all 3 will apply, but i think (3) deserves more attention when it's received. when almost anyone can build and release a monetizable app, video, or music album just by iterating with AI tools (not possible just yet, but soon!), resource distribution becomes much more random—and, in one sense, egalitarian
@talyarkoni Regarding 3): we're already there with random "influencer" success on YT or TT. The probability of success is very low, though, and will be even lower if LLMs increase the pool of people who can implement "good ideas".
@twitskeptic agree that influencing is basically already this. but i don't see why P(success) should be lower once the same dynamics apply to creative output. i would expect people to consume _more_ AI-generated media, not less ((b/c the quality will be better). and since the models are the great equalizers, and ~anyone can use them, it seems more likely to me that we'll see an endless stream of 1-hit wonders than a very small number of massive successes
actually, i think there are two things going on here: the denominator will presumably get bigger (because of lower barriers to entry), which i think is your point. otoh, the gains will probably be distributed more uniformly, b/c talent and skill will presumably play a smaller role. so i guess i can see a variety of potential outcomes, ranging from "everyone who does this for a while will eventually get 15 minutes of fame" to "only the people running media botnets and SEO ops will win"
@talyarkoni I'm guessing that available attention is already saturated, so I think more AI generated media will just lower the value of most media in general. And there will be a huge knockoff problem if it's cheap and fast to create a non-identical but similar version of a big media hit. So I'd guess your 2nd choice will account for the majority of successes.
But who knows? These big technological shifts are always so unpredictable.
@twitskeptic agreed. this also assumes the notion of a creative class continues to retain any meaning, which it may not. plausibly, the future of media consumption is just a set of cheap subscription services that let each of us generate our own never-ending feed of one-off works tailored to our own idiosyncratic tastes. MidJourney already feels that way—i doubt a lot of people are hanging _other_ people's MJ creations on their wall