That satisfaction gap between executives and workers reads like the same mental model problem, one level up. Executives aren’t using the tool every day so they might miss the hallucinations, the rework, the false confidence. Workers see it.
I agree we can all learn to use AI tools better, and I think there’s also a real risk of overconfidence as our knowledge increases. Confirmation bias is real and also extremely enticing.
AI literacy isn’t just about strong prompting and understanding the tools better. It’s knowing how to verify and ratify what we’re being told, especially when the information we receive sounds polished and convincing.
Yes, specially verification is often problematic. Output from AI sounds very good also when it's wrong. That's why there is an Appendix in my course with the title "Reading AI Numbers Without Being Fooled" :)
Matija, I read your article through a vendor-side lens I've built over 25 years selling advanced technologies into enterprises. Your broken-mental-model diagnosis is right. Let me share what I've watched happen in every disruptive-technology cycle — orchestration, cloud, SaaS, now AI: the mental model gets broken before the ramp-up period has had time to complete. The reason sits underneath technology and underneath process. Every disruptive tool touches the core human need to feel safe and pushes back against the discomfort that new knowledge brings, because it questions the intelligence built over time around the previous one. The buyer's quiet question is: why remove something that has worked so well? When daily practices are disturbed, skills are questioned, and people's relevance is put at risk, the reaction is to retreat to the reference point where safety lives and the old models protect. The ones most exposed resist most. The smart vendor holds the ramp-up open long enough for a new reference point to form, without forcing actions that break the mental model prematurely. Whether this is doable in a business world where short-term gains matter most is the real question. Perhaps AI itself will reveal smarter ways to hold that window open.
I am one of those “too generic guys” so I built my own problems, solved them and learned how to use AI effectively. So the way you explained totally makes sense here..
Absolutely agree. Also, I think starting with the small, repetitive tasks got me in the mindset to be looking for them. Then the productivity compounds. Great read, thank you!
That satisfaction gap between executives and workers reads like the same mental model problem, one level up. Executives aren’t using the tool every day so they might miss the hallucinations, the rework, the false confidence. Workers see it.
Yes, the usual outcome is that managers don't use it and workers don't have enough training
I agree we can all learn to use AI tools better, and I think there’s also a real risk of overconfidence as our knowledge increases. Confirmation bias is real and also extremely enticing.
AI literacy isn’t just about strong prompting and understanding the tools better. It’s knowing how to verify and ratify what we’re being told, especially when the information we receive sounds polished and convincing.
Yes, specially verification is often problematic. Output from AI sounds very good also when it's wrong. That's why there is an Appendix in my course with the title "Reading AI Numbers Without Being Fooled" :)
Matija, I read your article through a vendor-side lens I've built over 25 years selling advanced technologies into enterprises. Your broken-mental-model diagnosis is right. Let me share what I've watched happen in every disruptive-technology cycle — orchestration, cloud, SaaS, now AI: the mental model gets broken before the ramp-up period has had time to complete. The reason sits underneath technology and underneath process. Every disruptive tool touches the core human need to feel safe and pushes back against the discomfort that new knowledge brings, because it questions the intelligence built over time around the previous one. The buyer's quiet question is: why remove something that has worked so well? When daily practices are disturbed, skills are questioned, and people's relevance is put at risk, the reaction is to retreat to the reference point where safety lives and the old models protect. The ones most exposed resist most. The smart vendor holds the ramp-up open long enough for a new reference point to form, without forcing actions that break the mental model prematurely. Whether this is doable in a business world where short-term gains matter most is the real question. Perhaps AI itself will reveal smarter ways to hold that window open.
Thank you for this great insight!
I am one of those “too generic guys” so I built my own problems, solved them and learned how to use AI effectively. So the way you explained totally makes sense here..
Absolutely agree. Also, I think starting with the small, repetitive tasks got me in the mindset to be looking for them. Then the productivity compounds. Great read, thank you!