I recently read Herman Melville’s novel entitled The Confidence-Man. (At least I assumed it was a novel when I picked it up, although often it seemed more like a series of anecdotes.  I need to reread it.)  In the story, Melville portrays a series of con men (or is it just one?), pretending to be a physician, philanthropist, pharmacist, philosopher, psychologist, investment advisor, etc.

One of Melville’s characters observes that it is almost impossible for the economy to function without trust:

Confidence is the indispensable basis of all sorts of business transactions.  Without it, commerce between man and man, as between country and country, would, like a watch, run down and stop.

Unfortunately, evolution gives an advantage to those who take advantage of others.  Fortunately, an equilibrium is reached where only a small part of the population consists of sociopaths.  Here’s Melville:

“Pray, which do you think are most, knaves or fools?”

“Having met with few or none of either, I hardly think I am competent to answer.”

“I will answer for you.  Fools are most.”

“Why do you think so?”

“For the same reason that I think oats are numerically more than horses.  Don’t knaves munch up fools just as horses do oats?”

One character argues that only machines can be trusted:

“I’m now on the road to get me made some sort of machine to do my work.  Machines for me.  My cider-mill—does that ever steal my cider?  My mowing-machine—does that ever lay a-bed mornings? My corn-husker—does that ever give me insolence? No: Cider-mill, mowing-machine, corn-husker—all faithfully attend to their business.  Disinterested, too; no board, no wages; yet doing good all their lives long; shining examples that virtue is its own reward—the only practical Christians I know.”

“Oh dear, dear, dear, dear!”

And he anticipates the day when machines shall replace all workers:

“Hence these thousand new inventions—carding machines, horse-shoe machines, tunnel-boring machines, reaping machines, apple-paring machines, boot-blacking machines, sewing machines, shaving machines, run-of-errand machines, dumb-waiter machines, and the Lord-only-knows-what machines; all of which announce the era when that refractory animal, the working or serving man, shall be a buried bygone, a superseded fossil.”  

Melville wrote this in 1857, relatively early in the new industrial economy.  In some ways, the issue of “confidence” is probably much more central to today’s economy than to the economy of the 1850s, when most Americans were farmers and lived in small villages where people knew each other.  The following is from a 2021 paper on psychopathology:

In this direction, it has been proposed that it would be possible to find higher levels of psychopathic traits in certain professions or occupations (e.g., entrepreneurs, managers, politicians, investors, salesmen, surgeons, lawyers, telemarketing employees). The reason behind this could be that it is precisely these traits the ones that could boost the tasks involved in those professions or occupations and even facilitate success in them (Hare, 2003b; Dutton, 2012; Babiak and Hare, 2019; Fritzon et al., 2020). . . . 

In this regard, Dutton (2012), after applying the Levenson Self-Report Psychopathy Scale (LSRP; Levenson et al., 1995) online to 5,400 people and asking about their profession, found that, in the United Kingdom, the 10 professions with the highest levels of psychopathic traits were company CEOs, lawyers, radio or television characters, salespersons, surgeons, journalists, priests, police officers, chefs, and civil servants.

There’s a great deal of discussion about what AI will or will not be able to do in the future.  I see less discussion of how AI will address the problem of confidence.  Who would you trust more, a human selling herbal remedies, or an AI dispensing advice on herbal medicine?  A human real estate agent, or an AI providing information on houses for sale?  A human investment advisor or an AI giving investment advice? 

If Melville is right, then the biggest thing that AIs will have going for them is that people trust machines more than they trust other people.  Recall how Uber mostly solved the taxi regulation problem.  Perhaps AIs will solve the principal-agent problem.

PS.  This is a link to Melville’s novel.  Chapter 7 provides an amusing take on effective altruism (especially the portion from the bottom of page 53 to page 58.)