I feel like what a lot of these “AI is not a bubble” articles get wrong is that they approach the topic from a flawed epistemic frame.
Im not saying that I know AI is a bubble. I’m nobody and I know nothing. But I don’t think any financial crisis looks like the previous ones because those risks are the ones everyone is looking out for.
I think “how do bubbles pop” is the wrong question in general and the question should hardly be financial, the real question is:
“How do systems fail?”
What measures that aren’t though to move together move together?
Where are personal incentives misaligned from global incentives?
What else is feeding off the same feedback loops?
I think when trillions of dollars are involved, you have to start from
Skepticism. As they say in poker if you can’t identify the rube, you’re the rube.
Fraud is too strong a word, but if you can’t find the folks risking systemic contagion to make a buck, look harder. Maybe it’s contained! But start from the assumption we missed something.
But even the way this is worded maybe hides how narrowly scoped this article is:
“A shock to the expected returns of AI poses little danger of a recession along the lines of 2008.”
Essentially 1 of thousands of possibilities is unlikely. This says nothing about all the others.
I feel like what a lot of these “AI is not a bubble” articles get wrong is that they approach the topic from a flawed epistemic frame.
Im not saying that I know AI is a bubble. I’m nobody and I know nothing. But I don’t think any financial crisis looks like the previous ones because those risks are the ones everyone is looking out for.
I think “how do bubbles pop” is the wrong question in general and the question should hardly be financial, the real question is:
“How do systems fail?”
What measures that aren’t though to move together move together?
Where are personal incentives misaligned from global incentives?
What else is feeding off the same feedback loops?
I think when trillions of dollars are involved, you have to start from Skepticism. As they say in poker if you can’t identify the rube, you’re the rube.
Fraud is too strong a word, but if you can’t find the folks risking systemic contagion to make a buck, look harder. Maybe it’s contained! But start from the assumption we missed something.
But even the way this is worded maybe hides how narrowly scoped this article is:
“A shock to the expected returns of AI poses little danger of a recession along the lines of 2008.”
Essentially 1 of thousands of possibilities is unlikely. This says nothing about all the others.