It might be the beginning of the end… Though that article is really, really bad. It’s framing the loss as “AI is about to take over, people think software is going to be worthless because AI will write it and make all the software companies go bankrupt”.
The situation is much, much more volatile than that, but you need to look at the bigger picture. We’re not just talking 400 billion. We’re talking up to a trillion, depending on how you interpret the data.
There was some loss in software companies unrelated to AI, yes. But most of the loss comes from companies investing heavily into AI. This is because the market is getting more and more nervous.
During the big hype, everybody needed some investment in some AI company in case it pays off. The hype now seems to be over. Anthropic has released a feature that manages to slightly differentiate it from the competition. It’s not a complete game changer or giant leap forward or anything, it’s just that they may have finally found a use case it could actually be useful for. Word is about automating legal analysis, marketing and customer service tasks. Note that none of these are directly software-related. Claude is still a very, very comfortable distance away from writing usable software without humans actually doing the design.
The trouble here is that the enthusiasm has somewhat dimmed after all major companies made insane investments into AI the last 2 quarters. It is getting clear that not all of that is coming back. What is happening here is not so much people panicking because software companies are going to be made worthless by AI. If anything, software companies should become worth more if they can boost their productivity and reduce spending on manpower, right?
But none of that is the issue here. The issue is that Anthropic is threatening to take a clear lead. People are no longer betting on any random horse. it’s no longer too early to tell who will win the race. People are now worried for their investments and try to bet on the winner, selling off shares of companies that don’t look like they’re in a position to win the AI race anymore.
The question of whether any company will win that race is still as open as it ever was. Possibly more so now that I expect to see more and more culling, and more and more nervous market behaviour reacting to any hints that one company might be getting an advantage over the others.
It’s gonna be a nervous year. And quite frankly, they’re moving faster than I thought. But I’m really not that great at predicting economic timescales. They keep surprising me both ways…