Maybe we are not doomed by AI

Last week, I read some terrifying articles in which two people shared their perspectives of where all this AI shit is headed.

Cal Newport, however, does not agree with Matt Shumer's take on AI, his argument being that the job-takeover of AI is much more narrow than we are being told, with articles like Shumer's being more AI doom-clickbait and "vibe nonsense." He believes that disruption is going to be very limited to computer programmers, and very specific areas of technology.

I don't think Newport is wrong. AI is undoubtedly a powerful tool, as was Google when it came out. When I started my career in the early 2000s, Google was still new. Using Google was synonymous with being advanced, working fast. It didn't replace people, but it helped people do their jobs faster. It also created a clear divide between people who were able to adopt a new technology (web search, broadly) and those who refused. It changed how we work, but the work was still there.

The more I use AI, it feels the same. Yes, it can probably be scaled to replace some jobs - basic coding, really basic research, etc. But anything beyond those tasks is going to require humans. Anything even within those tasks will still need a human to oversee and make nuanced decisions. AI may speed up a coder's work, but will it replace the coder entirely?

Much of this remains to be seen of course, and I can't predict the future. But I can look to the past to inform what I see coming. I also spend way too much time online and it has become clear that "AI DOOM CLICKBAIT" sells, so anyone talking or writing about how the world will end because of AI, is doing it in part to make money, not necessarily to accurately inform.