Full Disclosure: Hydrazine Capital, an investment fund lead by Sam Altman (CEO of OpenAI) is the lead investor for my employer Geo Land Solutions, which specializes in real estate valuation to help municipal property tax assessors perform more accurate, fair, and equitable valuations.

I think it’s fair to say that people, including myself, have consistently underestimated the pace and capabilities of AI technology. It’s entirely possible that we’ll start to run up against some sorts of physical or theoretical limits soon, but even if that hits us tomorrow, the amount of raw power that’s been unlocked in the last year alone – the tech that’s already here – already has the potential to change things in massive waves.

In the coming year I’d like to explore a lot of these different scenarios. I’m not necessarily super confident in any one of my predictions coming to pass, but I think it’s important to explore and consider various possibilities.

I’m a strange creature in that I’m equal parts techno-optimist and neo-luddite depending on which day of the week it is. I think AI has the potential to massively enrich and empower the human race, but it also has some deeply troubling implications. Today’s post paints one of those troubling possible futures.

Summary:

A Market for Lemons
In which the internet gets clogged with piles of semi-intelligent spam, breaking the default assumption that the “person” you’re talking to is human.

The Great Logging Off
In which human biological and cultural evolution responds to the new super drugs we’ve unleashed upon ourselves for better and for worse.

First I need to bring in a market failure concept from economics called a “Market for Lemons.” Let’s say you have a used car you’d like to sell. It’s a pretty good car. But the problem is, you can’t prove it’s good. For all the buyer knows, your car is actually a lemon–a used car that seems fine on the day it’s purchased, but is actually going to break down in horrible ways days later, after which you are long gone with the money and the buyer is screwed. As it so happen, your actually good used car–a plum–is not a lemon, but because you can’t prove it’s not a lemon, you can’t charge the price a plum should command. This sucks for anyone trying to sell a plum.

The average selling price of any used car in the market depends on the ratio of plums to lemons because every buyer’s price is discounted by the risk of getting stuck with a lemon. So long as there’s even a few lemons in the market, the average market price for a used car will always be slightly lower than what a genuine plum would go for in a world where where buyers could accurately distinguish them.

This means sellers of plums start to take their merchandise off the open market. If you can’t sell your plum for what it’s actually worth, you might as well drive it yourself until it’s dead, give it to a family member, or sell it off the open market to a friend who knows and trusts you. The problem is, now there’s one less plum on the market, which increases every buyer’s risk of purchasing a lemon, which means the average selling price of all used cars goes down, which means more plums leave the market, and the vicious cycle continues until you’ve got nothing left but a “market for lemons.”

The surest sign of a market for lemons is when everything for sale is garbage, and nobody has an incentive to put something up for sale that’s not garbage, because they have no way to credibly signal their offerings aren’t garbage, too.

To a certain extent this has already been happening to many parts of our great global telecommunications network. For instance, when was the last time you answered a phone call from an unknown number at a time you weren’t expecting a call? Back in the early 1990’s it would never occur to me not to answer