AI and the individual human

Ted Gioia on the fate of creators; Dwarkesh Patel on replicating middle managers; David Friedman on self-employment

Ted Gioia writes,

Creating itself feels like a sucker’s game for losers.

That should never be true, but especially not now when distribution happens via software, without warehouses or delivery trucks or any of the traditional expenses associated with moving product.

If I make a song, and you listen to it, how much should the intermediaries between us earn for their distribution software? If you ask me, they should be getting the pennies, while the creators get the dollars. But now, it’s the other way around.

I wish that I could quote Gioia without picking on him. But he is engaging in adolescent whining about Corporate America. Ironically, this meets the market test, because he has orders of magnitude more readers than I do.

Look, the fact is that digital technology has drastically lowered the cost for creators to produce music or videos and to make them available. So supply has increased.

If the intermediaries were the cruel oppressors that Gioia portrays them as being, then the supply of creators would be going down. But instead, more and more people are making music and videos, so supply is exploding.

I will grant that the supply of movies that I used to go to movie theaters to see may be going down, but that is because consumer demand has shifted. The mass market is not there any more. For better or worse, most people prefer to be entertained by what is on their phones than what is in the theaters.

In entertainment today, the hard part now is not creating content or making it available. The hard part is getting attention. Intermediaries, like Netflix and Spotify, make money by bringing creative works to the attention of consumers. But the intermediaries are also in a precarious position, because the supply of content is growing faster than the available attention.

Consumers want to pay attention mostly to what other consumers pay attention to. This means that creators live in a winners-take-most world. Call it the Girardian creator economy.

No intermediary is going to be able to change it so that a large fraction of creators can be winners. It’s not the intermediaries’ fault. Get. Over. It.

Dwarkesh Patel writes,

All of Google’s 30,000 middle managers can be replaced with AI Sundar copies. Copies of AI Sundar can craft every product’s strategy, review every pull request, answer every customer service message, and handle all negotiations - everything flowing from a single coherent vision.

Pointer from Tyler Cowen. This is an interesting scenario to noodle on, but I have to admit I have a hard time seeing it on the near-term horizon. The same with the next scenario:

David Friedman writes,

As AI gets better, more and more of a firm’s workers can be automated. Eventually there might be only one.

He is raising the possibility that as an entrepreneur with the help of AI you can do everything yourself. If you do need help assembling your product or service, Friedman points out that

Improved communications, improved search technologies, reduce the transaction cost of coordination through the market, shifting the balance between market and hierarchy.

He concludes,

There will still be firms in the future, some employing thousands of people; I do not expect technological changes to produce a economy where everyone is self-employed.

But that is the direction we are moving.

A world of small artisans, using the Internet and AI to market their goods and services, may sound idyllic. But be careful what you wish for.

The outcome for independent entrepreneurs in the economy of the future may resemble the outcome that Gioia is observing in the “creator economy” today. Intermediary firms like Meta and Google, having made large investments that pay off in establishing powerful networks, will earn high returns. Individual artisans and entrepreneurs will compete in a winners-take-most lottery, in which the vast majority will not earn enough to be able to quit their day jobs.

This future, if it arrives at all, seems to me to be a long way off. Looking at AI relative to the current state of the job market is an exercise that will lower your expectations for rapid disruption.

Also, note that both Patel and Friedman site Ronald Coase’s theory of the firm, which says that a the choice between outsourcing a function and going it in-house depends on the relative transaction costs involved in the two approaches. But Friedman thinks that firms will get smaller, because of what AI can do for one person interacting with the market, while Patel thinks that firms will get larger, because of what multiple AIs can do within a single firm. I suppose you could suggest that they are both right, and we will end up with large single-person firms. But I think this goes to show you how unclear and speculative the future is at this point.

substacks referenced above:

@

What fully automated firms will look like

@

Shrinking Hierarchy

@

"The Infrastructure of the Recording Industry Is About to Fail”