Motivational Characteristics of Marketplace Participants

I have thought about issues related to “Motivational Characteristics of Marketplace Participants” for quite some time already. Although that title is already quite long, I think I need to narrow down what I intend to write about today even more.

I want to contrast two instances of marketplaces — one from the past, the other from the here and now. They are in fact quite similar — and some might even argue that one has been replaced by the other — namely it almost seems as if “Yellow Pages” have been replaced by “Google”. Both of these institutions seem to resemble natural monopolies. Yet they are also quite different in some respects: whereas “Yellow Pages” (and similar business directory organizations) made money by listing companies in their directories, “Google” (and other brands considered to be so-called “search engines”) usually make money by collecting and selling information about their users (in most cases — but not exclusively — to companies who pay Google to target these users with targeted advertising).

One very significant difference between traditional business directories and so-called “search engines” like Google … which exists on a very fundamental level is that while Google is a “one size fits all” offer, business directories tend to be more focused, specialized and specific. This can be a matter of geographic localization, or it can be a specialization in particular market segments or a wide variety of aspects. Such specializations are often also reflected in the languages of the people who participate in the corresponding communities — whether that is a local dialect, or a professional jargon or some other kind of language variation.

In contrast, machines and machine learning systems like Google attempt to apply simple algorithms to each and every case, regardless of variations that could be meaningful in specific cases. For example, many people can easily and obviously observe such simplistic (or oversimplistic) solutions when they see an algorithmic translation of text regarding a subject in which they have some specialized knowledge.

Ironically, there seems to be a lot of interest in so-called “large language models” in the field of so-called “artificial intelligence”. Over time, I expect the main insight real people might gain from such algorithms is that machines can create a whole lot of text simply by combining sentences like “see spot run” and “there was a run on the market” to create more (and even longer) press releases no human being with a rational brain would ever waste a single second to look at.

Then again human beings are indeed not very interesting for companies like Google. Their main motivation for business is to maximize profit, and as long as they are able to find enough suckers who are willing to give them more money for useless crap, then making more useless crap will continue to be a winning business strategy.