better check those bibliographies Gemini provides if you’re having it write research papers. It hallucinated well over half the sources AND didn’t provide the correct date in one citation. and that happened just last week. DON’T trust but verify - and with a human librarian. thank goodness i did! it has also cited a public policy that doesn’t exist and i would have used that to make decisions had i not caught it in time. my concern is that too many of us are trusting it too much too soon.
The most dangerous question in the AI age sounds pragmatic:
“Where do we put the humans when AI does the work?”
It isn’t pragmatic. It’s diagnostic.
It reveals a hidden premise: humans are primarily functions—and “not being needed” becomes an existential defect.
AI doesn’t just change jobs. It relocates the bottleneck: from output to judgment—criteria formation, accountability, refusal, and the ability to stay coherent under uncertainty.
![[ Center for Humane Technology ]](https://substackcdn.com/image/fetch/$s_!uhgK!,w_120,h_120,c_fill,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5f9f5ef8-865a-4eb3-b23e-c8dfdc8401d2_518x518.png)
better check those bibliographies Gemini provides if you’re having it write research papers. It hallucinated well over half the sources AND didn’t provide the correct date in one citation. and that happened just last week. DON’T trust but verify - and with a human librarian. thank goodness i did! it has also cited a public policy that doesn’t exist and i would have used that to make decisions had i not caught it in time. my concern is that too many of us are trusting it too much too soon.
Thank you for this informative post!
Appreciate the way you frame today’s unease about AI and work
—> the uncertainty around both skills and identity is something many people feel.
It aligns with a deeper cultural pattern I unpack in From Frontier to Server Hall, where I explore how work’s role as a social
anchor shifts as AI scales:
https://open.substack.com/pub/saraeson/p/from-frontier-to-server-hall-moral?r=1t83gn&utm_medium=ios&shareImageVariant=overlay
The most dangerous question in the AI age sounds pragmatic:
“Where do we put the humans when AI does the work?”
It isn’t pragmatic. It’s diagnostic.
It reveals a hidden premise: humans are primarily functions—and “not being needed” becomes an existential defect.
AI doesn’t just change jobs. It relocates the bottleneck: from output to judgment—criteria formation, accountability, refusal, and the ability to stay coherent under uncertainty.
Essay here:
👉 https://open.substack.com/pub/leontsvasmansapiognosis/p/the-most-dangerous-question-in-the
— Leon Tsvasman