34 Comments
User's avatar
Richard Jordan's avatar

You wrote: "If the hour estimates from 1500 are remotely accurate, and if behavior then was roughly adaptive, then we’d need the value we get from fiction to be far larger today, to justify spending eight times as many hours on it." I think this fails to compare MARGINAL values. A 16th-century Medieval peasant could substantially improve the welfare of himself, his family, and his community by spending an extra hour of scarce daylight working; spending an hour consuming fiction came at a heavy cost. What is the cost now to spending an extra hour of one's evening watching tv instead of sleeping? Electric lights are cheap (unlike candles) and tv is cheap (unlike plays). The value of one fiction-hour in 2025 is probably much less than in 1525, but it might still be rational to consume more of it, both from the standpoint of the individual and of the species. I think you are on stronger ground arguing that our fiction is maladaptive, than arguing our consumption levels are.

Expand full comment
Jack Crovitz's avatar

Fun article! But I’d take issue with this key sentence:

“If the hour estimates from 1500 are remotely accurate, and if behavior then was roughly adaptive, then we’d need the value we get from fiction to be far larger today, to justify spending eight times as many hours on it.”

I see no reason why the amount of fiction consumed in 1500 would be adaptive, because fiction in 1500 was subject to immense supply constraints. In that world of subsistence farming, there were simply not many people with the time to produce fiction. Also, they lacked the technology to effectively distribute fiction (printing technology existed but was still extremely expensive). Because the supply of fiction was so constrained, consumption of fiction was surely less than optimal!

Consider, by analogy, the consumption of antibiotics in 1500. Because antibiotics were so supply-constrained in that premodern age, it would be ridiculous to assume “consumption of antibiotics in 1500 was roughly optimal.” It would be even sillier to say that we should return to 1500-era consumption of antibiotics!

That said, this is an interesting question and (for what it’s worth) I’m not sure we are at optimal fiction-consumption nowadays.

Expand full comment
TGGP's avatar

Antibiotics didn't exist in the 1500s, and we can see the obvious benefits we get from them now, which don't actually take up that much of our collective time.

Expand full comment
Jack Crovitz's avatar

Agree that antibiotics is not the best analogy to use here.

Perhaps protein consumption would be a better analogy? Many people in 1500 did not consume the optimal amount of protein due to supply constraints! It would therefore be very silly to say we should revert toward a 1500s level of protein consumption *because* of an assumption that 1500 had a roughly optimal consumption level. The exact same logic applies to the argument Professor Hanson is using here with respect to fiction consumption.

Expand full comment
TGGP's avatar

Not the exact logic. Consuming protein takes less time than consuming fiction.

Expand full comment
Tim Tyler's avatar

The point is that assuming consumption levels in 1500 are anywhere near optimal has no obvious logical or empirical basis.

Expand full comment
Dominic Ignatius's avatar

Like sugar and porn, fiction/leisure/entertainment is probably something along the lines of stuff was supposed to point us in the right direction, and maybe serve as a little treat now and then, but has potentially become a maladaptive super-stimulus in modern times. Probably will be as hard to combat as things like sugar and porn, too. Good luck!

Expand full comment
Phil Getts's avatar

I recently got to try out the Apple Vision Pro VR headset. Visually, it's very hard, perhaps impossible, to distinguish the VR from reality, except for things within a few feet of you. I think it literally has one pixel per retinal cone. If game developers get their act together, and use AI to control interactive characters and to manage exciting personalized plots, it will be extremely addictive.

Expand full comment
Berder's avatar

I can't believe nobody has said this: ChatGPT is not a legitimate source of data on which to base an analysis!

Expand full comment
Robin Hanson's avatar

"legitimate"? By what rules, set by whom?

Expand full comment
Berder's avatar

This should not need to be explained. ChatGPT makes up whatever numbers it imagines might follow, with no relation to ground truth. It is no better than just making up whatever numbers you think sound good. (No, that's not a source of historical data either!)

Expand full comment
Lupis42's avatar

I buy the notion that the modern amount is above the average optimum, but the idea that the 1500 amount is a useful baseline is absurd.

The modern world requires a lot more frequent social interactions with previously unknown people where the social rules are not clearly spelled out - someone living in a small town today can expect more even than life in the biggest cities in 1500. Fiction is one of the major ways people expand their models of other minds, and game out ways such interactions could go.

Expand full comment
Ben's avatar

Is adaptiveness the right lens to look at how people spend their time these days? Surely the ultimate goal of humanity would be something closer to "all humans living happy lives". Reading a book isn't very productive, but it's probably a lot more enjoyable than working all day in a farm or factory.

So it makes sense that the amount of time spent on fiction (be it books, movies, games, whatever) would increase, because people genuinely enjoy these things. This might be a sign that we're getting closer to a society where people prioritize happiness over productivity. And imho that's a good thing.

This assumes that people are actually genuinely enjoying their time with fiction however. Many do, but not everyone. There are many people who deep down feel numb/empty/bored as they lay on the couch all night, and that doesn't seem ideal from either a happiness or produtivity point of view. But that's a separate issue.

Expand full comment
Robin Hanson's avatar

Prioritizing happiness over productivity is probably maladaptive, and so cultures who embrace that will go away, to replaced by more adaptive culture.

Expand full comment
Phil Getts's avatar

I don't think this is "the answer", but it is something to consider: The marginal adaptive value of fiction should be higher in a more-complex world, and in a world with more social mobility. If your only options are "continue to be a serf" or "become a brigand", you don't need to do a lot of scenario gaming.

Expand full comment
Robin Hanson's avatar

A bit more, but 8x more?

Expand full comment
Phil Getts's avatar

Like I said, it's just one factor to consider. (Also, as you know, 8x as much consumption doesn't translate into 8x as much value.) My gut feeling is that the amount and the type of fiction we consume are both maladaptive.

Expand full comment
Alexey Romanov's avatar

> By averaging estimates of ChatGPT 3, 4.5

Isn't this strictly worse than using ChatGPT 4.5, or o3, etc.?

> then we’d need the value we get from fiction to be far larger today

No, we'd need 8 times as many people to be able to read at all. Based on the table from https://en.wikipedia.org/wiki/Literacy#Medieval_and_early_modern_eras, this is actually an underestimate even for Europe; so a typical _literate_ person would be reading less fiction than in 1500!

Expand full comment
Alexey Romanov's avatar

To be fair, illiterate people could listen to poetry.

Expand full comment
Tim Tyler's avatar

I concluded that modern entertainment is net negative here: https://timtyler.org/destroy_the_entertainment_industry/

Expand full comment
Gesild's avatar

Could it be that consuming modern fiction contributes to collecting skills, wealth, allies, power, kids, something?

Expand full comment
Zvi Mowshowitz's avatar

One could decompose this change into leisure consumption over time, and the share of that consumption that is fiction. It seems plausible that given the decision to consume so much leisure, it is not a mistake to consume this much fiction, or it is a much smaller mistake. So the focus should be on our potentially maladaptive increase in total leisure.

Expand full comment
Catherine Caldwell-Harris's avatar

Hours a day consuming fiction is not adaptive, as you note. The explanation is that cultural innovations hijack evolved adaptive systems without fulfilling (all of) their adaptive function. Steven Pinker wrote about fiction in this regard in the 1990s.

... Funny, I asked Google AI, and AI disagreed with me, claiming: Reading fiction doesn't "hijack" humans' evolved systems in a negative or harmful way, but rather utilizes and enhances existing cognitive mechanisms that support our ability to understand and interact with the world and others. AI cited: https://pmc.ncbi.nlm.nih.gov/articles/PMC4733342/

Expand full comment
Catherine Caldwell-Harris's avatar

I do understand, from reading this blog for a while, that Hansen may concede that there are evolutionarily-based reasons why humans watch fiction 7 a day. But this is irrelevant to his point. As he says in one of the comments, since this isn't adaptive, our society should change this and other non-adaptive habits, because if we don’t, our culture will be replaced by a more adaptive culture.

With globalization, cultures that advocate a strong work ethic succeed and increase their population share and territory occupied. China is a good example; the overseas Chinese have been spectacularly successful (same for other Confucian-values countries).

Is massive media consumption the modern bread and circus? Elites need to keep the masses occupied and somewhat content so they don't try for more power-sharing.

Expand full comment
Catherine Caldwell-Harris's avatar

I see Dominic Ignatius and perhaps some others in the comments mentioned the hijacking/super normal stimulus perspective.

Expand full comment
Nicholas.Wilkinson's avatar

Lions do.

Expand full comment
Vaughn Svendsen's avatar

Seems that we should all become preppers.

If things don't fall apart, then we don't need to learn new skills.

If things do fall apart, then learning prepping skills is highly adaptive.

Expand full comment
John Ketchum's avatar

According to your figures, Americans spend an average of nearly 7 hours a day consuming fiction and music, and more than three hours a day on each. I assume that's for all Americans, regardless of age or employment status. Is it possible that young people and the unemployed spend a disproportionate amount of this time? I'm also curious as to what counts as fiction. Does that include TV watching?

Expand full comment
James Hudson's avatar

We no longer need to worry much about adaptation: we have achieved such mastery that now we can mold the environment to suit ourselves.

Expand full comment
Robin Hanson's avatar

Not remotely true. Our cultures are drifting into maladaption, for which we will pay big prices.

Expand full comment
James Hudson's avatar

One lives in a house, drives a car over paved roads that take him where he wants to go, works in an office, dines in a restaurant, takes medicine when ill, etc. All these are designed to fit him—he does not have to fit them. Adapting to our circumstances is less important when we have created those circumstances for our convenience.

Expand full comment
Robin Hanson's avatar

We just don't know how to, nor are we in the habit of, remaking our cultures to suit us for the long run.

Expand full comment
James Hudson's avatar

The *long* run—yes, that’s impossibly difficult. But, of course, we never have achieved long-run adaptation, except by luck. And as for the short, and even the medium, run—we now achieve that more by modifying our environment than by changing ourselves.

Expand full comment