America is gloomy about the economy despite the numbers. That's smart.
I was still an investment banker before the financial crisis sent me running for something stable like being a novelist. It wasn’t a great fit – I was the only guy on the trading floor who could work a bow-tie. Yet there I was, arguing with the head trader about his baffling faith in the “quants.” He said that A) They were smarter than me, and B) had access that I didn’t have, and C) in case I’d missed his first point, they were smarter than me. Who was I to argue? I pointed out that the social context – ego, status, pride, terror – was not being run through those models, and in retail real estate, context matters. He graciously reiterated points A & C, and told me to “shut the fuck up.”
As it was, the quants made the world more liquid than rich and we used all that credit to buy a global financial meltdown.
These days the two parallel stories of today's uncertain economy are “Look how well the economy is doing!” and “Why are we so gloomy about it?” The argument is that there is a disconnect between perception and reality. There always is, but I’d argue that it isn’t that people aren’t facing reality, but what is being counted in the data is entirely different to what is being calculated in people’s heads. Data reflect past quantifiable reality well, but it doesn’t do much to perception, which moves reality going forward.
These two related but different sets of calculations have led to the widespread belief that something is wrong, despite being told otherwise. We can’t quite put our finger on it, but we are terrified that when we do it will be an unmitigated disaster.
Inflation is high. The core consumer price index may strip out volatile food and energy costs, but consumers can’t practically strip those expenses out of the budget. Interest rates are sapping credit personally and nationally; the US debt service is now higher that we are paying for defense which, by the way, is going to have to increase if we don't want to find ourselves facing down the entire world. And we might.
Wars are looming in Europe, the Middle East and Asia. The Sahel in Africa is in one of its coup-o-rama phases. That might seem remote, and it is, but on September 10th 2001, the average American couldn’t find Afghanistan on a map. Haiti is a collapsed state in our backyard pool. The states of Latin American aren’t at war (yet), non-state gangs have made the place the most violent place on earth. Those migrants and refugees aren’t walking to Buenos Aries to take part in Argentina’s libertarian experiment. We know which way they’re heading.
Four years ago we had a summer of unhinged summer protests that were capped off by a Trump v. Biden presidential election. It ended badly. We’re about to have another. Leaving the rest of the Republic to sort out the nagging dilemma: “What next, how much is it going to hurt and what will it cost me?”
Well, business runs in cycles; war is all hell but they eventually end, but this AI foolishness is a different sort of uncertainty. Like the quants whose models failed to see the financial crisis, AI will scrub enormous data sets to deliver an extruded, liquified deli ham-loaf of conclusions with unwavering confidence.
If you happen to be a human, you could be forgiven for being a little paranoid, despite the numbers. Human heuristic calculations don’t process nearly as much information as a computer and never will. There is a certain advantage in running an operating system that has evolved to understand that you don’t have all the information. We do have the odd advantage of knowing when we don’t know something, which enables us to take a blind-spot into consideration, even if we can’t see it.
To put it in Rumsfeldian terms: Humans are aware of the “known knowns”, as well as the “known unknowns”; and the smart ones are aware of the existence of the “unknown unkowns.” Data models consider two thirds of what you need to consider if you plan to stay alive. It is precisely this lack of complete information that dominates human decision making and perception. We don’t really care if the collected data says everything is rosy, we know damned good and well the model has missed a spot. Even if we haven’t got a clue what that spot is.
I mean, just look what happened with the internet…
Comments