Automated trading can magnify errors and weaken human oversight

Transcript

From the transcript How I Invest publishes with the episode.

1. 16:15

   How do you think you know, that's an example of sort of, you know, this automated trading and financial market players giving some control to machines, letting them make decisions in that millisecond scale where people can't compete. So the world has financial world has already given some autonomy to machines. At what point do they really get uncomfortable with that, with AI agents and, like, trying to automate?

2. 16:44

   Well, tell you one thing that I think is that is worrisome is in the olden days, like twenty years ago, you could have a room and by the way, I'm using analogies from trading, but it applies universally, where it's something where you'd have the intuition of knowing whether something was right or wrong. You could have a cacophony of noise all around every you know, the person next to you is fighting with his wife, someone's, you know, arguing with her husband and blah blah blah.

3. 17:13

   Million things. But if somebody said something wrong, a wrong price, or somebody was buying and should have been selling, or someone was saying the wrong thing about a company, the room would stop and everybody would focus because it was an intuition about things. Now everything is in a glop in a box. And the things are going so fast and analyzing so much information that you you can't into it anymore.

4. 17:41

   You it spits out an answer. What are supposed to do with it? You can't criticize it. You have to accept it. And that's where there's an opportune you know, leverage goes both ways. The age of this kind of thing, you can have you couldn't have a a mistake that would cost $10,000,000. You know? You could do things backwards or you know? Well, how could it cost you that much money? Today, you can have an errant piece of software that in one minute exercises 100,000 transactions, and you could lose billions of dollars.

5. 18:14

   In other words, the world leverages made things better, but it also makes things more concentrated and dangerous and take away the intuition element. One of the things I like about your product is, you know, let me go things I'm more familiar with. The thing about Google was always, it's a bibliography. You get the sources and you have to investigate it. Then you come up with AI and it gives you the answers.

6. 18:40

   But that's not totally satisfying because how do I know to trust it? What I want is I want the answer with a bibliography. I want to trust but verify. I want to get the thoughtful, you know, intelligent response to my question and maybe an elaboration of it. But I want to go back and look at the sources. Or at least verify that there were sources. And I think that that's an important element going forward of all these different devices.

7. 19:10

   Mere bibliographies aren't going to work. And the answer is not going to work. Because who could trust? You can't trust it. And you don't wanna trust it. And you don't want a world where everyone's gonna just trust it because then why do they need anybody? No.