this is a wonderful essay. it addresses a myriad of ways in which our human efforts at understanding complex systems go awry. I particularly appreciate the essays discussion of the need to be symmetric in our thinking. Why is something right as well as why is something wrong in any given hypothesis? I also deeply appreciate the authors discussion of the symmetry as I see it between errors in human thinking and what is called AI hallucination. I’m deeply appreciative of the effort that went into writing this essay. It’s very valuable for me. Thank you.
I am a “child” of the 1960s, which, I suppose, makes me suspect…, but it certainly makes me old and shopworn! There were then some nationally divisive issues that led to one of the recurring opposition mantras from that turbulent time: “the personal is the political.”
If you have “nut jobs” in responsible positions, you may get bad policy, or even dysfunction systems. In fact, it is hard for me to see how we might get sound social, cultural or political orgaizational and decision-making systems with people who have character disorders like borderline personality or narcissism. Does not dysfunction start with people and then morph to their organizations?
Maybe, and good luck with this, we might litmus test political candidates before political parties nominate them. Maybe, we might even have more than a binary political party system. If there’s one thing both parties can seem to agree on it’s “let’s keep it at two!”
Thank you for a deep insight into the seam that measures human/AI accepted patterns and the hole where variables meet using explicit and tacit posits that completely distort reality.
A rare positive proposal for a way forward. We have many explanations of the problems, so few plausible solutions. Thank you.
I can't resist a question. If AI provides "boundary intelligence" that then feeds into human judgment, who has the authority to judge? That person would need a foundation of social authority. This is not likely to be something that experts or institutions would welcome:
"The expert closes around a lane. The model closes around its corpus. The institution closes around its procedures. The guild closes around esteem. Each protects its inside."
The Fall of Icarus. At first I thought you were referring to Braudel's Structures of Everyday Life (it's on the cover) until I looked up the painting. The institutions have indeed flown too high, exceeded their proper places. That leaves a wasteland of legitimacy. Rational institutions (the original AIs) and capital systematically dismantled the alternative bases of authority that once existed. Who is left?
I am coming back to read this when I get time, but just yesterday a customer said he wanted the updated EU REACH SVHC statement, and gave us a precise date of 3 June 2026.
Really? News to us.
No offense to Boomers or anyone older than 25, but because of AI, old ways of keyword search are out the frigging window with AI.
Unless there is an alert from the governing body who makes these changes, don't bother looking for anything out there. You should be on their mailing list already for update alerts. If not, it is the wild west out there for regulatory.
this is a wonderful essay. it addresses a myriad of ways in which our human efforts at understanding complex systems go awry. I particularly appreciate the essays discussion of the need to be symmetric in our thinking. Why is something right as well as why is something wrong in any given hypothesis? I also deeply appreciate the authors discussion of the symmetry as I see it between errors in human thinking and what is called AI hallucination. I’m deeply appreciative of the effort that went into writing this essay. It’s very valuable for me. Thank you.
I saw this Icarus painting in Vienna!
I am a “child” of the 1960s, which, I suppose, makes me suspect…, but it certainly makes me old and shopworn! There were then some nationally divisive issues that led to one of the recurring opposition mantras from that turbulent time: “the personal is the political.”
If you have “nut jobs” in responsible positions, you may get bad policy, or even dysfunction systems. In fact, it is hard for me to see how we might get sound social, cultural or political orgaizational and decision-making systems with people who have character disorders like borderline personality or narcissism. Does not dysfunction start with people and then morph to their organizations?
Maybe, and good luck with this, we might litmus test political candidates before political parties nominate them. Maybe, we might even have more than a binary political party system. If there’s one thing both parties can seem to agree on it’s “let’s keep it at two!”
"Institutional hallucination" should be a new term in common parlance. Fascinating.
Brilliant essay thank you Craig for sharing these insights 👍
Thank you for a deep insight into the seam that measures human/AI accepted patterns and the hole where variables meet using explicit and tacit posits that completely distort reality.
Your article deserves further study.
A rare positive proposal for a way forward. We have many explanations of the problems, so few plausible solutions. Thank you.
I can't resist a question. If AI provides "boundary intelligence" that then feeds into human judgment, who has the authority to judge? That person would need a foundation of social authority. This is not likely to be something that experts or institutions would welcome:
"The expert closes around a lane. The model closes around its corpus. The institution closes around its procedures. The guild closes around esteem. Each protects its inside."
The Fall of Icarus. At first I thought you were referring to Braudel's Structures of Everyday Life (it's on the cover) until I looked up the painting. The institutions have indeed flown too high, exceeded their proper places. That leaves a wasteland of legitimacy. Rational institutions (the original AIs) and capital systematically dismantled the alternative bases of authority that once existed. Who is left?
I am coming back to read this when I get time, but just yesterday a customer said he wanted the updated EU REACH SVHC statement, and gave us a precise date of 3 June 2026.
Really? News to us.
No offense to Boomers or anyone older than 25, but because of AI, old ways of keyword search are out the frigging window with AI.
Unless there is an alert from the governing body who makes these changes, don't bother looking for anything out there. You should be on their mailing list already for update alerts. If not, it is the wild west out there for regulatory.