From my perspective biggest shift is that people that would never have made anything digital 10 years ago are now making their own apps and websites. It is cool to see because often they make things that currently do not exist. The downside is that we software people that have been in the trenches for a while also learn about laws, cyber security and other topics that you should know, but claude won’t tell you unless you know what to ask for. These people have never heard about any of these things, and what I have seen so far getting built ranges from slightly risky to straight up super illegal. All without no clue what they are doing, no warning from claude, and the timeline from zero to launching something illegal can now be done by the average person in an evening.

This is both good and bad. We get more new stuff, people enjoy creating these websites, but they risk themselves and others people in the process. Like a kid being allowed to drive a partly autonomous car.

The super annoying thing I have encountered are managers and other low technical people with some influence over engineers that now believe they ”know the limitations of AI” because they managed to make a CRUD demo in 30min. More than once have someone in my vicinity expected claude to one shot an algorithm that might be physically impossible to even do. I have also seen people try to ”prove” that it is possible by having claude make this algorithm on some toy example where the output looks okay at first glance, but with no verification that it is actually correct. And now the engineers ”just need to take it to production”. AI have made any professional software development 100x worse in my experience. Annoying people that previously could be convinced that certain things are not quick and easy are now extra stubborn, extra annoying, extra time consuming to deal with. Therefore, in my professional life, AI is a big net negative when it comes to efficient use of time. AI made all bad things worse, and barely brought anything good with it. I am convinced that it doesn’t have to be like this, but most organizations are dysfunctional to some level, and that just became worse. If I could go back in time, I would not have gone into software engineering. I love the field, but there is just too much BS going on.

I also see and hear about tiny teams being more effective than ever, teams that already have the skills to know what they are building, and how. They can ship more, faster, and still ensure reliability because they know what they are doing, and they are allowed to spend the time if needed to ensure things are done correctly (without AI powered ”managers” standing in their way).

I think we will see many small companies literally come out of nowhere, no funding, no big names behind them and completely crush companies 1000x their size because of AI. I also believe big slow organizations will use AI to become less efficient and ship worse products while panicking on their way down, ensuring ”all in on AI” by sending more emails faster than ever.

More easy problems will be solved than ever, every imaginable easy problem will be possible for anyone to solve. The hard novel problems will remain unsolved, remain difficult and only the smartest, most stubborn, most stupid people will attempt to solve them, and if they do, the market will reward them.