Personally, I feel that as an individual, it's the right time to complete a program, but as a team, it's become harder.

It's true that the proportion of founders has increased both in the US and in my country, Korea.

And unlike the old days, it feels like what's needed now isn't so much deep, concentrated programming knowledge in one area, but rather broad knowledge across many fields. The claim that "productivity has increased" really only applies to freelancers. In fact, there's been a noticeable increase in freelance outsourcing requests that would be hard to handle without AI, lots of short deadline gigs compared to before. And of course, that makes it harder to charge appropriately.

For teams, on the other hand, you still need things like code reviews and team decision making.

As an individual, I've practically become someone who just writes up a gate, lets AI handle the code, checks that the core domain doesn't break, watches the gate's rules, and pulls the lever.

The reason team work slows down is mainly because Agile methodologies and code review processes are still human centric and consensus driven, and human cognitive speed itself becomes the bottleneck.

So I can understand a lot of the arguments that come up in the comments. The important thing is that most people tend to only see their own situation and their own context, which makes it hard for them to understand others.