The gradual numbing effect of AI adoption

Recent conversations with friends and peers who have been proactively using AI have surfaced their fears of not really knowing or understanding what they have up to with work. They seem to have the results but with litle understanding of how and why. One of them, a younger newer entrant to the workforce, said that he had stopped learning anything new since using AI. Another brought up a bigger concerns, something I sensed from time to time over the past year. He felt a complete lack of interest in wanting to try anything new. The motivation just wasn’t there anymore. He tried blaming it on age at first, but AI seems the more likely culprit.

In my case, it’s that odd impression given by those who have never been hobbyists, pass comments claiming they can do what I could, but in minutes with AI. With any task just moments away from completion, there isn’t a strong need to attempt anything. Avoiding AI for me, has kept that flame alive as I continue to try new things at home and get better at them. Andy Balaam explains this better than I can in his video and blog post.

On the subject of getting things done in moments, I challenged a work peer very excited about the infinite possibilities of AI. I asked him to list out recent, life-changing products that have come into existence only because of the presence of AI. Quicker search results, and AI itself, were the only things. When the internet became available to us, I could list 10 - 20 things then possible. So, is slop content, better searches and the hype of being able to create anything the best AI can do? GitHub says monthly commits went up from 1.4 to 2.9 billion since April of 2026. What is the impact of this? A colleague was excited that ChatGPT agreed with him on a recent purchase. But when I probed, and laid out a few more circumstances and factors - things I understood from years of knowing him. Would ChatGPT still be a good judge?

Entry barriers are inherently elitist, but somehow software written by LLM-aided developers and art designers makes me skeptical. I second-doubt every open-source project now, every piece of art and that takes energy. It ruins the joy of discovery and exploration. The ethics of how LLMs operate are dubious, and I want no part of it. Even F-Droid is now swarmed by AI-generated apps. A blog post where the author studied about a 100 or so recently updated apps on F-droid found that nearly 80 per cent of them were LLM-generated/aided apps. This is not good for open-source and people. For me, it’s another layer of checking.

Every time I see a post about someone who ‘built’ a new tool or created art with just prompts, this seems like people riding the hype train, something put nicely by Veronica Explains in one of her recent videos. Somehow, I no longer attribute the product’s success to its developer. This is a stark difference in my attitude, compared to a year or two ago. If AI allows you to make something cool and wow, in moments, I do not need you to do it for me. If I want, I’ll generate it myself.

Some things are more than skin-deep. It’s annoyed me to see AI slop show up when I was updating some Linux packages using yay. I uninstalled the package promptly, but I am not sure how long I’ll be able to keep this up as the floodgates to the Linux kernel and Debian have been opened for LLMs.