The New York Times had a solid retrospective on Zuckerberg's insistence that the metaverse was the future. His company has mostly put that effort on life support. I mean, did anyone not see that coming? No one really wanted that. No one really trusted a company that knew it was turning teens into addicts. And a company that required people to work in the office obviously wasn't that sold on virtual work itself.
What's interesting to me is how so many business leaders bought into it, even though they couldn't really define what it was. Media companies in particular seemed particularly nervous about it (I'm looking at you, Disney!). But I'm not here to celebrate more job losses, or offer "I told you so." I'm more interested in looking at the hype cycle, and what the next one looks like.
The suggestion that the metaverse was something we needed for meeting virtually of course seems silly when we already learned to work via Zoom by then. This time the hype is about AI, and it's different. It's different because AI, in certain contexts, is useful. It's the context that is often misunderstood. We've already seen an unearned confidence in AI tools that have caused leaders to cut back on their workforces. They don't have much to show for it yet, and what little academic study that has been produced suggests it's not going to work out the way they expected.
Think about it. LLM's are just really impressive algorithms that can identify patterns. And while they're programmed to exhibit what looks like confidence, it's just a semantic choice. LLM's do not innovate, they are not creative, and above all, they can not read your mind. The optimists seem to believe that LLM's are capable of all of the above.
What they are outstanding at is generating code under supervision from experienced engineers, quickly, reducing the length of the iteration cycle. That allows stakeholders of all kinds to get to the product they want faster. The opportunity is not to do more with less, it's to do more with more. This is the adjustment that needs to be made in this hype cycle. AI is, fortunately, not the metaverse.
No comments yet.