A bold statement, I know. People tend to believe an answer or advice from AI sooner, faster, and more completely than one from humans. That is a major danger.
I just saw an informative post by Anton de Nijs in which he argues that GPT5.2 scores 145 on an IQ test, making it smarter than the smartest employee at most companies. And that the smartest will always win. (My very short summary (1)).
My response to that was that your smartest employee usually isn’t immediately proven right; rather, the opposite is true. People have a built-in distrust of those who are (much) smarter. If people with high IQs were believed immediately, six hours of meetings could be condensed into 15 minutes (2).
The smartest people in the room have to spend months, if not years, massaging and talking before their outcome is accepted by the organization. This is often when things have already gone wrong, just as the smartest person predicted beforehand. That person is not celebrated, but they were seen to have always been against it, so that must be the reason why it went wrong… (Smarties; choose your battles…)
How differently people deal with AI. Although AIs regularly hallucinate (GPT 5, about 50% of the time!), it is immediately accepted in most places as manna from heaven. No further questions, checks, second opinions, or other opinions seem necessary.
The implicit underlying assumption is that smart people will have suspicious personal goals and that AI is completely neutral. This is a misconception about both.
The problem of (hidden) bias in AI is well-known in the IT industry. A small error in the training data can quickly spiral out of control, causing the AI to produce incorrect answers or discriminate against certain groups. (3)
Gifted individuals are not only extremely smart and fast, but often also have a very strong sense of justice (4). While it may seem otherwise, those are precisely the people who are not after your job. They simply want “what is best” for the organization and/or the customers.
If we all keep chasing AI as the source of wisdom and ignore the truly smart people in the room, this smart person predicts that we will make very wrong decisions that could turn out very badly for humanity. (5)
But you don’t have to take my word for it; take your time with these links.
(1) post van Anton de Nijs
(2) hoogbegaafd op kantoor
(3) AI bias
(4) rechtvaardigheidsgevoel
(5) Rampscenario
