OK, but what if an Amazon algorithm has actually learned that people who search for "shirt without stripes" are more likely to buy more things if the first image they see is a picture of a striped shirt?
The original version of which talks of an "Invisible hand" and if that's not an metaphor for divine intervention hidden by an chaotic system i don't know what is.
It's rather surprising how often almost all complex systems theories be it AI, cosmology or economics have an aspects where even the theorists are resorting "to because it is".
Sometimes those statements are based on measured data but it's not always easy or possible to do so accurately for highly interconnected system or worse system where you have actors reacting to theoretical model in a way that changes how the system behaves.
in that case the applied AI stops being a Search tool (as was the purpose of the search bar) and becomes a new Ad tool. And this masquerading is not a great thing at all, for the same reason why people don't like bots pretending being humans during phone calls.
But it’s not a search tool. It’s a make Amazon money tool, as all their tools are. I think you misunderstand why amazon build these tools if you think they are to make your life easier in trying to locate things to buy on their site. That’s a happy coincidence. They build them to make money.
I don't think it violates the efficient markets hypothesis. I think that entails "if it was possible for someone to create an AI that would make more money by showing something else, the AI would be showing something else." Roughly, efficiency means that the most profitable solution will be the one that wins out, but "profitability" includes the costs of building such an advanced system, assuming that it's possible at all.
I think it's more likely that they've A/B tested faulty algorithms and picked the one that's the least faulty.
I don't have proof, but I strongly believe that a search algorithm that returns what a customer is actually searching for will drive more sales. I suppose it's possible that with time, consistently bad results will beat a customer into submission and drive more sales of stuff the customer doesn't want. But I don't believe that's true, and this would only be the case if the customer accepts that the thing they want doesn't exist. If the customer is pretty sure that solid color shirts exist, they'll just shop elsewhere until they find it.
It's a tricky balance between sales and relevance. If you don't watch for relevance you will end up showing only booze and porn in the commerce search results, border-line-porn in video search (true story), and so on.
Then the algorithm would not be acting in the customer's interests.
Presumably this would be after the algo devalued people who clicked on "Next Page" until they came to a page that had stripeless shirts on it, or who, after the search, only ever clicked on stripeless shirts. "Deeds not words," dontchaknow.
