More Thoughts on AI Adoption

Stub • 418 Words • Artificial Intelligence, 2026 • 03/27/2026

⚠️ This post contains a rougher cut of my thoughts on the topic and may be updated in the future. Please forgive any mistakes or lack of polish!

The various facets of when we do/do not want to adopt AI and why this happens.

There are 454 words in this article, and it will probably take you less than 3 minutes to read it.

This article was published 2026-03-27 00:00:00 -0400, which makes this post and me old when I published it.

AI, Isolation, and Collaboration

  • “Having a place to say “something feels weird … can you help me figure this out?” at any time feels deeply useful, particularly for the “somethings” that feel itchy but not yet important” (Cosleuth by Robin Sloan)
    • Your other friends, co-workers, etc don’t have your context or are busy with their own work that having a broadly learned chat partner always around gives you a low friction way to bounce ideas around and iterate
  • AI sycophancy can drive you away from people who criticize you (@spns.tv)
    • Could it make you less immune to being challenged by others? Increasing defensiveness, etc.
  • This overall can drive more isolation which in turn makes you rely on the chat platform even more

AI and the human desire to be special

  • “i think something in the divine awaits me for resisting”
    • i don’t want to be like other people
    • i am smarter/better/more capable than other people
    • i am more moral than other people
  • All in all I think there is a kind of moral smugness in a lot of anti-AI sentiment, which they do have some justification for.
  • “the ai will never be as smart as humans”
    • so what?
    • are what you always using all of your effort and intelligence? the law of large numbers and statistics can’t approximate doing the same thing over again with slight parameters tweaked?
    • we are smart but not as smart as we like to think

AI, Toil, and Willingness

  • AI can take the toil and dirty work away from you
  • this increases my willingness to take on work
  • it does also unblock some tasks like reading PDFs
    • it is so hard to do this programmatically, but there is a lot of bureaucratic data trapped in these kinds of files
    • if there are tons of PDFs you don’t want a person doing all that repetitive, mind-numbing work

Shipping Speed with AI

  • So where are all the AI apps?
  • Taste and well-defined roadmaps are the most important part because if implementation costs go down then what becomes more important is what to implement
  • There also has to be safeguards against the torrent of slop
    • There are some people who may have too much trust in AI agents and become “slop cannons”
      • i won’t even call it AI brain rot but rather an acceleration of the Dunning Kruger effect like people will overestimate their own ability and underestimate their stupidity via AI
    • People and understanding of AI output then becomes the bottleneck
Click here to see a list of all the external links cited in this post

You might also like...

Thoughts and Questions Around AI Adoption

I tend to be cautiously optimistic about AI, so I wanted to put together the kinds of things I'm curious about around AI, particularly adoption.

AI Usage Acceptability Spectrum

A write-up of my own intuitions about what (potentially) makes certain usages of AI more acceptable than others.

You Won't Get Left Behind

On AI hype and why you can just watch it ride out for now.

AI Art and the Intention of Differentiation

Introspecting on the difference between AI and human art, specifically in taking input and producing output.

Biomimicry and AI

My notes on Dwarkesh Patel's interview of Andrej Karpathy.


Other Posts About “Artificial Intelligence

Gourmet Software

A new food-based metaphor about a potentially emergent category of software.

Thoughts and Questions Around AI Adoption

I tend to be cautiously optimistic about AI, so I wanted to put together the kinds of things I'm curious about around AI, particularly adoption.

AI Usage Acceptability Spectrum

A write-up of my own intuitions about what (potentially) makes certain usages of AI more acceptable than others.


Comments