Post #3394776

2026-06-24 23:57 UTC

The AI 2027 guy is taking credit for being a good forecaster: www.lesswrong.com/…/daniel-kokotajlo-s-shortform?… Recently an interviewer asked me how I got to be such a good forecaster, and I replied by saying something humble. In retrospect it was a bad answer because I should have instead used the opportunity to give actual advice on how to forecast AI well. Reminder, according to this guy (originally, he has already nudged the dates of his prophecies back a few years), we are supposed to be having agents replace human workers in mass this year, and be one year out from a superhuman coding agent and researcher that radically iteratively improves itself. We joked about how the AI 2027 people would try to pivot as they were proven wrong… I think they are going to outright try to claim credit for being ‘right’ despite all their critical prophecies being wrong by claiming credit for secondary details.

Replies (2)

@YourNetworkIsHaunted@awful.systems 2026-06-25 01:24

It’s also funny to note that at least 2/5 of the points are actively bad advice. Naively extrapolating from a trend line is one of the most common errors people make when trying to make a prediction, especially when you’re already prone to letting the aesthetic of data lead you astray. Trusting in a kind of “normalcy bias” or whatever you want to call the assumption that the world will continue to be pretty normal is one of the better ways to hedge against that. Also I’ve said it before but the name “technological singularity” literally comes from the idea that at the hypothetical rate of change they’re posting all our existing models of what is possible or probable break down like the laws of physics at the center of a black hole. If you’re reasoning from a pre-singularity model then definitionally there is no expectation that it should continue to hold true. I don’t think I need to get too deep into why the whole singularitarian concept is pretty sketchy in its own right, but since it still lies at the heart of the science fiction driving these people’s predictions I think its worth acknowledging that it does suggest its own nonsense.

Open ##3395117
@lurker@awful.systems 2026-06-25 01:36

They are writing another scenario called “AI 2030” (which we already discussed here) so I wouldn’t be surprised if thats the angle they play

Open ##3395155