Do you use AI to expand your agency or surrender it?
I have always loved the way @wesmckinn thinks and writes. My Data Science career started glued to his books.
I wholeheartedly agree that loops for laziness' sake surrender agency; what he is calling agentic engineering expands it.
I have often experienced 'clanker' behavior from LLMs and agree that 'agent' is too grandiose a term for current capabilities. I use similar principles in my projects, but I have some great ideas from this post that I am excited to try out!
https://t.co/9k5oyNLJwX
Source: Steinacker-Olsztyn, Gosain & Dao, Is Misinformation More Open? A Study of robots.txt Gatekeeping on the Web, ACM Web Conference 2026. https://t.co/D0WvmlKaEg
Does AI have access to the same internet you do?
The credible web may be opting out of AI faster than the junk web.
A 2026 ACM study:
60% of reputable news sites blocked ≥1 AI crawler.
Misinformation sites: 9.1%.
What does that do to the evidence AI gets to reason over?
AI is getting incredibly easy to build;
trust is getting harder.
I asked a vendor how I could trust their AI. They said: “Other companies use it.”
That’s social proof, not verification.
The real AI moat is proving your product reliably does the job a customer needs it to do.
Will we let AI expand or erode our Agency?
As AI’s capabilities increase it seems like people increasingly surrender their Agency in the spirit of progress.
We can choose to lazily surrender our agency, or use AI to bring the biggest increase in agency in human history.
Which means evals, observability, provenance, human escalation, and failure recovery aren’t infrastructure around the product.
They’re part of the product.
The real moat in AI isn’t the model. It’s trust.
People use AI products hoping they’ll actually solve the problem. But LLMs come with empirical performance, not theoretical guarantees.
The winning products won’t just generate.
They’ll verify.
Check out my latest article: Building on Unsettled Waters: Why Automated Testing is Crucial for Generative AI Applications https://t.co/hXim8gR6hj via @LinkedIn
I used to think “I’m not racist”, but my time in St. Louis, especially during the Ferguson protests, really opened my eyes. I used to think I was empathetic to the inequality that was so prevalent among minorities in our country.
https://t.co/Ql5i4kprhP
If I take this viewpoint, I am more likely to be open to engage in conversation of those affected (hopefully a much more amicable conversation than is Lucas’ song). I believe a big part of the problem is many of us don’t recognize the problem in ourselves.
You’ve got to be kidding me!! Vice President of the United States Mike Pence said Domo!! I didn’t see this happening. So stoked that all of my team’s work and all their all-nighters are paying off. Huge shoutout to partners Qualtrics and Nomi Health too. Wow. @RyanQualtrics