Prediction: The Growth of the Vibe Sector
Where are the bottlenecks when you add intelligence and stir? Part 1.
Many more things will be valuable precisely because they are not AI generated. That is the prediction—the growth of the Vibe Sector.
As a quick review, consider three areas that went from making decisions on “mostly vibes” to “just a little bit on vibes”: baseball, fashion, options trading.
The old-timey scouts were not onboard when sabermetrics swept through sports analytics. Competition played out and they were replaced by the nerds. “The guy’s an athlete.” “Good jaw.” “He’s got an ugly girlfriend, which means no confidence.” All those attributes that the scouts used to know based on vibes were quantified, ranked, and turned into WAR—wins above replacement, the rent above opportunity cost. This turns out to be way more effective.
Scouts for fashion models too have lost out when the vibes of a “look” is immediately quantifiable by Instagram numbers. I don’t know if there is WAR for this.
Doing the hard work of being a genius and creating a better options pricing model lets you mop up on the trading floor—Myron Scholes (edited slightly):
That was in 1973. Then, it was the case that there was the old grizzly traders who thought they had the experience from the over the counter market and the new young turks who were going to be market makers and trade on the floor of the Chicago Board Options Exchange.
So here's an idea with experience only and intuition—versus a model. The young guys had the model. Fisher Black made sheets of paper which talked about the Delta and the pricing at different levels of the stock price relative to the exercise price. And they could look at the sheets.
And there was a war between the grizzly intuition people and the model people, the young turks who had no intuition, but they had the model. And in a matter of about six months or so, the young turks had wiped out the grizzlies, okay, the intuition people.
Cause intuition is a model, but it's an incomplete description of reality. All our intuition is a model; you can do it, but it has a big error to it. So if you could use a model like the Black Scholes model, it reduces the error and set up an ability to hedge at that time.
There will always be an error term when modeling the world. Setting up the model takes intelligence, sometimes Nobel caliber. The people estimating the error term to get an edge after using the model, let’s call them the vibe sector.
If you add intelligence and stir—that is, give everyone access to superhuman level intelligent machines (ASI)—this will certainly give us way better models of the world in lots of areas. But, and here is the kicker, many goods are valued precisely because of error term. It isn’t always the case that there is an objectively true value that the model is estimating; the true value is endogenous to the estimation, if you will. Lots of problems are private value, not common value auctions. The vibes—and the people who judge the vibes—matter. That is what generates the value for the product.
It isn’t that some ML algorithm cannot digest an options exchange (or Instagram) and calibrate a slightly better options pricing (or fashion) model—this I assume is already being done. It is that consumers will value certain things because it has the attribute “decided by a human, not an AI,” and this will become much more important over time.
People might now be okay with lab-grown diamonds (is it because divorce laws and marriage norms decrease the need for marital credible commitment mechanisms?). But some people still like to pay a premium to know their thing was Made in the USA. Why? Because vibes. Just like life insurance was taboo because it mixed market and moral affairs, we will erect taboos around the human sector. Because vibes.
To be more specific is to be more speculative, but here goes: we will want human educators, human friends and nannies, human religious leaders and mentors, human travel agents and exercise trainers, human sales people, human book and blog writers. For these, people will say “Ew, you used AI for that?”
And precisely because these are valued for their vibes, we should expect intelligence to be labor augmenting instead of a labor substitute in these sectors. And only some of these are winner-take-all markets, so, yes, this should sound optimistic.