LLM AI programming agents are not good for mental health, it supercharges FOMO and takes procrastination to next level.
I have about 20 large changesets across 2 computers and 5 repos generated with AI agents ready to be pushed that I just can't force myself to review/take a look at.
Meanwhile there are small changes my apps need that I'm avoiding to do because I HAVE TO BE ON THE FUTURE TRAIN.
Meanwhile I'm constantly updating (and checking out what's new) 3 GUI frontends for AI agents, 5 code editors with agent plugins, opened about 50 tabs + about 5 new every day (that I'm afraid to look at) with github repos and search results for skills, MCPs and advice threads (I know these will become obsolete in about 2 weeks or so). I also sent myself some 30 emails with articles that I won't read.
I'm 54 and while I consider myself reasonably up-to-date (aside from Swift modern concurrency, it's just too much for me) this train feels just too fast for me, I kinda wish it rolled a bit slower (but then again, probably a lot of people felt this before).
Started with read-book-get-hooked-on-programming, in late 80ies had chance (from small rural city go to bigger city 2x/week) to use Электроника БК-0010 in programming club, 90ies-soviet IBM mainframe clone with Rexx in the uni & PC revolution (Turbo Pascal, love you) from 8080->Pentiums, co-made library software for universities (dBase and FoxPro), co-founded software dev company (Oracle DB&Forms, web, Lotus, 15 years), sold it, got burnout, first (original) iPhone and mobile apps ever since.
Now I'm looking on my business (my iOS and Android apps I worked on for more than 15 years) and then there's this chart:
And I know that @steipete and @mergesort and @dimillian are probably right and I applaud and respect @krzyzanowskim for being cool/skeptical with the train/hype and making https://CommanderAI.app (I hope you will sell it for billion), but it is exhausting even to watch. And I hate woods/handworking so that's also not an option.
Commander - The AI Workspace Beyond the IDE

An AI coding workspace for multi-agent development on Mac: plan, code, review, and release with clear diffs and worktree control.

Commander

@xjki @mergesort @dimillian @krzyzanowskim it’s okay to mourn our craft. https://nolanlawson.com/2026/02/07/we-mourn-our-craft/

it’s also okay to be incredibly excited about it and celebrate the fact that one person can now build an army.

We mourn our craft

I didn’t ask for this and neither did you. I didn’t ask for a robot to consume every blog post and piece of code I ever wrote and parrot it back so that some hack could make money off o…

Read the Tea Leaves
@steipete @xjki @mergesort @dimillian @krzyzanowskim there are not that many armies that need to be built and it’s difficult for any one person to afford their own health insurance, is why it is difficult to be excited and celebrate
@kyle @steipete @xjki @mergesort @dimillian @krzyzanowskim I am the happy retired continuous learner now. I still code because modern IDEs are way more fun than debugging machine code when I was a junior. I am inspired by all the recent posts/blog articles that reveal the LLM impact from many stages of people’s careers. This is more exciting than the web explosion, with different scary bits. I need to restart blogging to document this interesting time.
@stevehume I expect to never be able to retire, and to have to do manual labor when I am your age—if i am so lucky, on two accounts—specifically because of this technology. I hope you can understand why that is not an exciting prospect.
@kyle I do understand. Underemployed since dot com crash but health care is at least not a worry in Canada. My son, who works at Anthropic, thinks he has 2 years before the models are better than he is. Retiring is not actually a thing; not worrying about needing more money is. Then you just do what you want and the list of side projects keeps growing.