My hobby projects have 100x more tests than they used to, because LLMs are great at writing tests. And my subjective experience is that the net quality has increased as a result.
YMMV, but it’s certainly a common belief, and for me at least a lived experience.
A lot of people believe a lot of things are true. Plenty of things that are provably false too. People will have strong convictions. People will drink deadly Koolaid because they believe in things so much.
I don't care what people believe, I care about what is. What is measurable. What is factual. I need evidence for a belief to be meaningful. I need strong evidence for a belief to be strong. Not just evidence in favor, but evidence that alternative explanations are unlikely.
Currently I see evidence that things are moving fast. But I am unable to distinguish if this is actually because of AI or because increased efforts and motivation. Most importantly, speed isn't the same thing as velocity.
What I do not see evidence for is increasing quality. In fact, I see strong evidence to the contrary. I see strong evidence that quality is declining even quicker than it was before. I'm not convinced AI is that cause of this, but there's more than adequate evidence for me to believe it is a (significant) catalyst.
Right now you could throw a stone in a random direction and there's a good chance that whatever it lands on will be decreasing in quality. It is even easy to gesture broadly at Microsoft, but they aren't the only big tech disappointing their users.
I hear a lot of claims. I don't see a lot of evidence.
It’s fair to be skeptical. Though I’d question whether the quality declines you observe (presumably with hard evidence) really just started in the past 18 months or so, such that strong correlation to AI coding tools suggests causality.
For my part, I’m too busy building to spend time quantifying quality issues in my past decade of hobby projects so I can compare to my new work, which subjectively is far more capable and higher quality.
My hobby projects have 100x more tests than they used to, because LLMs are great at writing tests. And my subjective experience is that the net quality has increased as a result.
YMMV, but it’s certainly a common belief, and for me at least a lived experience.