Doing popular things

We come to contradictory conclusions.

Our tutoring consultant travels by bus out to work, or used to; he has now started to do remote tutoring, forced by COVID-19.  His fellow-travelers on the bus are, or were, mostly hourly workers, the kind who are most affected by this pandemic.  If they don’t work, they have no income.  Luckily for them, the government (and other entities) are coming up with ways to help them through this difficult time.

But he had a thought: suppose one of them had lost his or her job a year ago, through no fault but just because hourly work is like that.  Suppose a local drought stopped all landscape work, or some families had decided not to use nannies.  There was no movement to help such people then.  It pays to have the same problem as lots of other people.

It also pays to have the same likes and dislikes as other people.  Our photographer would like certain features in his cameras, but since few people agree with him, it’s not worth the camera company’s time and money to build them in.  Conversely, there are many features that he simply never uses, but other people do, so there they are.  This is less of a problem now than it used to be, since software is cheaper than hardware and most features are done with software.  But all digital cameras he can afford are still basically color and restricted to certain filters, rather than using a more flexible monochrome.

It’s true that the internet age has allowed some products to attain a critical mass of users by connecting a thin but far-flung set of potential customers.  We can cite the revival of instant-film photography, a technology of the 1960s and 1970s, for instance.  But consider an ever more important sector: machine learning, or Artificial Intelligence (AI).  This kind of software works by looking at a large mass of data and, behind the scenes, figuring out patterns.  These programs can be remarkably effective at picking out important items from a large mass of data.  But they rely on the working data being similar, in the mass, to the training data.  They only work by following the in-crowd.

On the other hand, the main effort of authorities these days is to ensure that many people are not getting sick with the same virus at the same time.  If you’re one of too many coming down with COVID-19 at once, there may not be capacity to treat you.  More mundane examples come to mind: driving in rush hour is not fun; a crowded beach in summer is less enjoyable.  And, of course, high-season flights and hotel rooms are more expensive.

We could add to our examples on both sides.  If you have a rare disease, there may not have been any effort to find a cure.  Our photographer would love to have certain films available that, alas, it is no longer economical to produce.  Your favorite band may have broken up because their audience wasn’t big enough.

Conversely, cameras or books for which there is little demand can be very cheap.  Paintings by an artist you like, but few others do, can be quite affordable.

So we conclude that marching to a different drummer is either a good idea or a bad one.