HomeArticlesHow do you outsmart the evil AI?

How do you outsmart the evil AI?

Kata Balogh 25 May 2022 1.7 min read Markets and systems
Öltönyös alak marionettzsinórokon, fölötte a bábvezérlő kereszt

AI eats data and breathes data; humans do not. It follows that AI can control humans but humans cannot control AI. Apparently.

Question: what does the AI know about you, reading this article at this very moment? Answer: a great deal - and practically nothing. All we have to do is find our way through that apparent contradiction, and we will have the key with which the evil artificial intelligence can be led astray.

Where AI is strong: the crowd

If AI has enough factual, relevant and precise data, it can tell you how any large community will react to some event. The forecast does not work for individuals and small communities, though: the reaction of a family is extremely hard to predict with any accuracy, because it decides and changes too quickly, and group dynamics barely come into play. With a hundred people you can already work with greater accuracy, and with a hundred thousand greater still.

AI evaluates data on the basis of references. If a scientific fact appears both in a tabloid and on a science portal, in case of a contradiction it will treat the second as valid. It classifies true and false according to what the source is, and how many sources confirm something.

You could say, then, that AI is infinitely deferential to authority.

The key against AI

If we put the statistical analysis and the reference-based classification together, we get the key to dealing with AI. The apex predator of the markets is AI (in reality a great many networked AIs), and as traders it is primarily against it that we are trying to win. The recipe: do not react in the usual way, and do not behave in the usual way. Do that, and we end up in the AI’s blind spot, booked as a rounding error in its calculations.

Explaining the essence of a strategy against AI is that simple, and understanding it is not much harder. Putting it into practice - that is genuinely difficult! Not impossible, only laborious.