Varieties of Deterministic Experience

What determines the fate of our world? Depending on your views, there may be few different macro narratives:

Mimetic Determinism

There are ideas floating around with various evolutionary properties. Some of them are really good at embedding in our minds, some are really good at making their hosts spread them further. The best memes get stuck in your brain, compel you to attain a position of great prestige, credibility and power, then insist that you spread them as wide and far as possible.

There are also meme-complexes (Memeplexes) with symbiotic or parasitic relationships. For example, some people are compelled to become Venture Capitalists, but it’s useless without the corresponding meme that compels people to become startup founders.

From Nadia’s The tyranny of ideas:

Rather than viewing people as agents of change, I think of them as intermediaries, voice boxes for some persistent idea-virus that’s seized upon them and is speaking through their corporeal form. You might think of this as “great prophet theory”.

Ideas ride us into battle like warhorses. We can witness, participate in, and even lead these battles, but their true meaning eludes us. We don’t really know where ideas come from, nor how to control them.

See Also
Wikipedia – Memeplex
Dawkins – The Selfish Gene
Joe Carlsmith – The innocent gene
Bernard Beckett – Genesis
Creanza et al. – Cultural evolutionary theory

Financial / Economic Determinism

There are incentives which necessitate certain consequences. If it’s profitable, it will be built. From Scott’s Meditations on Moloch

Just as you can look at an arid terrain and determine what shape a river will one day take by assuming water will obey gravity, so you can look at a civilization and determine what shape its institutions will one day take by assuming people will obey incentives… Just as the course of a river is latent in a terrain even before the first rain falls on it – so the existence of Caesar’s Palace was latent in neurobiology, economics, and regulatory regimes even before it existed. The entrepreneur who built it was just filling in the ghostly lines with real concrete.

More specific arguments can apply in local contexts. For example:

  • There are different dating apps.
  • Ironically, the ones that fail to create stable matchings will see more repeat use and popularity.
  • Thus, the dominant dating app will inevitably be designed to alienate.

More broadly, organizations may promote certain values, scientific institutions may promote the construction of certain types of knowledge. Platforms may promote certain kinds of content.

See Also
Wikipedia – Base and superstructure
Wikipedia – Economic Determinism
Applied Divinity Studies – How Substack Became Milquetoast

Game Theoretic Determinism

A perspective frequently utilized in super long term forecasting, and reasoning about alien civilizations or superintelligent agents. For example, an argument might take the form:

  • Some civilizations will colonize the universe
  • Thus, the far future universe will be populated mainly with agents who are pro-expansion and pro-progress

Or more mundanely:

  • Some religious groups on earth promote high-fertility rates
  • Thus, the near future will be populated mainly with agents who are pro-population growth

There are more nuanced and complex arguments of this general shape. In Liu Cixin’s Three Body Problem Trilogy:

  • Civilizations can’t trust each other
  • Technological progress can proceed exponentially, such that observing that a civilization is harmless from thousands of lightyears away is no guarantee that they’ll be harmless by the time you arrive
  • Thus, any civilization that learns about the existence of another civilization will immediately act forcefully to destroy it

See also:
Scott Alexander – The Hour I First Believed
Philip Trammell – Which World Gets Saved
Ben West – An Argument for Why the Future May Be Good

Physical Determinism

There are particles in the universe (including those in our brains) subject to physical laws which determine their behavior.


This is a limited typology, there are likely many other views, both reasonable and not. Note that these views are non-exhaustive. You could say that things are determined by physics at the level of individual particles, but by memes at the level of human behavior. Or that it’s all incentives, but that those incentives are modulated by facts about human psychology.

Note as well that these views should be neither horrifying nor necessarily comforting. To take another example, evolution does provide a guarantee that agents are well-adapted to some set of circumstances, but provides no guarantee that they remain well-adapted. Or in capitalism, there is a strong likelihood that sufficiently profitable business will be created by someone, but no guarantee that profit proxies well for human values.

Finally, these views all (perhaps with the exception of physical determinism) are not really absolute. You are perhaps a slave to memes as brain-parasites, but have some control over which memes you choose to take on. Through will-power may be subjugated to habit in general, and occasional breakthroughs of will-power might allow you to, for example, delete Twitter, sign up for therapy and install a Chrome extension that blocks content recommendations.

Identifying more of these views, mapping their influence in various social spheres, and better reasoning through their consequences is a promising avenue for future work. In particular, would love to see follow-up posts:

See Also
Katja Grace – What is going on in the world?
Yudkowsky – An Alien God
Yudkowsky – Inadequate Equilibria

Yes, X Causes Y

Holden wants to know: Does X cause Y? Whenever we try to figure it out, there’s an observational study, followed by maybe a couple more interesting studies that still don’t replicate, or have other horrendous drawbacks. To summarize the grand total of human knowledge of the subject: “it’s super unclear”.

It’s satire, but he’s also serious. Claiming inspiration from (among other sources), Scott Alexander’s “Much More Than You Wanted to Know“ series, Holden concludes “the bottom line usually has the kind of frustrating ambiguity seen in this conclusion.”

This is too pessimistic.

First, on the meta level, “does X cause Y” debates tend to occur right on the boundary line of epistemic confidence. So by nature the discussions are designed to be highly contentious and uncertain. We no longer ask “does X cause Y” questions about matters superseded by modern physics (“Does phlogiston cause combustion? No.”), nor about matters settled by empirical data (“Do germs cause disease? Yes.”). But this progress goes underappreciated.

For people living on the cutting edge of science, the important questions will always have ambiguous answers.

Second, object level, Holden is overstating the ambiguity of Scott’s evidence reviews.

  • In Melatonin, Scott pretty definitively concludes that Melatonin is an effective hypnotic, and the right dosage is 0.3mg.
  • In Autism, Scott concludes with ~100% confidence that “genes that increase risk of autism are disproportionately also genes that increase intelligence”.
  • In College Admissions, Scott concludes that “There is strong evidence for more competition for places at top colleges now than 10, 50, or 100 years ago.”

(I’m cherry picking a bit, but the point stands. In a good chunk of cases, we can be reasonably confident that X causes Y. I’m taking Scott’s confidence at face value here, but I think that’s reasonable, at least with regards to responding to Holden. In another post, Holden agrees that Scott’s predictions have a good track record.)

Third, most of the time, we care more about making the right decisions than about having the right epistemics. By which I mean, it’s okay to not be certain, as long as the expected value works out. For example, in Face Masks, Scott remains uncertain about the efficacy, but points out that for high-risk situations, the annoyance of wearing a mask is probably a price worth paying for even a small likelihood of reducing Covid risk.

Finally, it’s working taking a look at some really quick high level heuristics.

So look. Of course it sounds bad that 2/3rd of psychology studies don’t replicate. But that means that 1/3rd of them do!

It would be nice to know which third is which, but still, that’s not nothing.

Admittedly, it would be very bad if:

  • 1/3rd of studies replicate
  • 1/3rd show no effect
  • 1/3rd demonstrate the opposite effect

Though there is a bit of this in nutrition (“does X cause or prevent cancer? Maybe!”), it’s not universal. For example: I’m not 100% sure how well face masks prevent Covid, but I am pretty darn confident they don’t cause Covid.

This asymmetry matters a lot. And it matters for Holden’s more serious questions too. GiveWell isn’t totally sure if the parasite-killing drugs improve test scores, but they probably don’t make them worse.

So start with a reasonable prior (50% of all studies are wrong, or 66%, or 40% or whatever), update based on some object-level evidence, and multiply out to get an expected value estimate. As far as I can tell, this is immensely simplified, but basically what GiveWell does.

I’m not being critical of Holden. He knows all of this far far better than I do.

I’m just saying as a reader, when you encounter stuff like this, don’t throw your hands up in the air and conclude that since nutrition science is bad, we should all just eat pizza and whiskey for every meal. Don’t make the mistake of falling into epistemic nihilism.

To conclude:

  • Our discussions will tend to revolve around contentious issues, creating the illusion of an epistemic crisis.
  • Even on contentious issues, literature reviews are not useless.
  • The point is not to achieve certainty, the point is good decision making.
  • It might be possible to know some things some of the time.

[1] He clarifies this does not mean up to 40% of doctor’s opinions are wrong. Doctors are not relying on a single study. And again, I would point out that the bulk of doctor’s opinions are totally mundane and non-controversial.

Failures of Megaproject Management, or Why Are the Olympics so Unprofitable?

The Olympics are massively unprofitable for host cities, so why do they compete for the privilege of losing money?

To find answers, we’ll go on a meandering tour of how megaprojects go wrong.

Winner’s Curse?

To determine a host, each competing city submits a proposal, then active members of the IOC cast a vote, and the winner is chosen. Assuming more lavish proposals improve your odds of winning, this process is auction-like, and winners will tend to overpay.

That explains some degree of cost overrun, but what we see in practice is really tremendous. Baade and Matheson’s The Economics of the Olympics puts the London 2012 Olympics at a cost of $11.4 billion, with revenue at just $3.3 billion. Even if winners have a tendency to overpay due to the underlying mechanism, we’re still positing a situation where London makes an estimate, determines hosting is profitable, and then ends up being off by $8.1 billion. That seems like a fairly large error.

Indirect Benefits?

Okay, so direct revenue is low, but what about follow-on tourism, improved image, and so forth? In a section titled “The Long-Run Benefits of Hosting the Olympics”, Baade and Matheson examine the ongoing benefits of facilities, investment in general infrastructure, and future tourism, but ultimately conclude:

When bidding countries are appropriately compared with countries that are otherwise similar but did not bid for the Games using propensity matching techniques, the significant Olympic effects on trade, consumption, investment, and income all disappear. Again, the long-run benefits of hosting the Games prove to be elusive.

Although there may be some more ephemeral benefits, it seems unlikely that these sorts of indirect and long-run benefits make up for the immense direct financial losses.

Cities Don’t Compete?

In the past, my question would have had more force. In 2012, eight other countries competed with London to host the Olympics games. But these days, there seems to be less interest. For the latest round of selections, Brisbase was chosen “without a rival bid“. Perhaps cities are rationally responding to the profitability of the Olympics, and have ceased to express a serious interest in hosting.

Except that’s not quite the entire story. The AP reports that although Brisbase was the only city considered, it was not the only city interested. Doha and Budapest were both working on formal bids, as were cities in China, Germany, India, Indonesia and Russia. But all were excluded before they could complete the process, due to a last minute rules change.

So while it’s true that the number of cities issuing formal bids has decreased in recent years, there still appears to be widespread interest.

Cost Overruns?

Bent Flyvbjerg, author of The Oxford Handbook of Megaproject Management describes the Iron Law of Megaprojects: “Nine out of ten such projects have cost overruns. Overruns of up to 50 percent in real terms are common.” This bears out empirically for the Olympic games. According to one analysis, estimated costs are dramatically lower than actual, with Athens and Sochi seeing overruns of 400%!

This is certainly a piece of the puzzle, but it’s not the whole thing. Host cities have access to the same historical data. They know that their estimates are not going to be accurate, that the true costs will be higher, and can at least attempt to reason accordingly.

Maybe this is an extreme failure of inside-view thinking over reference class forecasting, or of extreme arrogance (“everyone else sucks at planning, but my estimates are better”), but that still feels insufficient. Lots of people have those problems in reasoning, they don’t go on to exceed budgets by 400% when tens of billions of dollars are at stake.

Intentional Underestimates?

The economist Albert O. Hirschman (of Exit, Voice, and Loyalty fame), also coined the principle of the Hiding Hand: if people knew the true cost of projects up front, they would never fund it. So insofar as our aim is to get anything done, some degree of ignorance is our friend. This bears out anecdotally, as when San Francisco mayor Willie Brown famously said

News that the Transbay Terminal is something like $300 million over budget should not come as a shock to anyone. We always knew the initial estimate was way under the real cost… the first budget is really just a down payment. If people knew the real cost from the start, nothing would ever be approved. The idea is to get going. Start digging a hole and make it so big, there’s no alternative to coming up with the money to fill it in.

It sounds crazy, but can be justified by appeal to rational irrationality. As Hirschman explains: “since we necessarily underestimate our creativity it is desirable that we underestimate to a roughly similar extent the difficulties of the tasks”.

Though Hirschman meant to celebrate the principle, Flyvbjerg and Cass Sunstein refute its glory, instead coining the “Malevolent Hiding Hand“. They note that empirically, underestimates of difficulty in fact greatly exceed underestimates of our abilities, leading to “unexpectedly high costs but also to unexpectedly low net benefits.”


If these overruns are not a case of rational irrationality, but rather of plain-old irrational irrationality, the question remains of how such important and financially impactful decisions end up being made by such incompetant actors.

The standard libertarian critique here is that there’s simply no market. It’s a top-down, centralized and bureaucratic decision, and therefore bound to fail. If the planners are wrong, there’s no way to “short” their position and correct the price, no mechanism that would improve a bad situation. As Yudwkosky describes an even more severe policy failure:

The Bank of Japan is just one committee, and it’s not possible for anyone else to step up and make a billion dollars in the course of correcting their error… to the extent the Bank of Japan has poor incentives or some other systematic dysfunction, their mistake can persist… Standard economic theory, generalized beyond the markets to other facets of society, did not seem to me to predict that the Bank of Japan must act wisely for the good of Japan. It would be no surprise if they were competent, but also not much of a surprise if they were incompetent.

Of course, you might hope that there are at least some personal drawbacks to losing $8.1 billion, but that doesn’t seem to be the case. Sebastian Coe, who masterminded London’s winning bid and chaired the London Organising Committee of the Olympic and Paralympic Games, seems to be doing alright for himself. Post-olympics, he was appointed Chairman of the British Olympic Association, presented with a lifetime achievement award from BBC Sport, and elected a member of the IOC.

So yes, incentives seem to play a role here, but we’re still not done. Given basic historical data, the reasonable thing for cities is to run away screaming as the prospects of becoming an Olympic host. It should never even get to the stage of an incompetant or poorly incentivized bureaucrat making a poor analysis.


From Baade and Matheson again:

economic concerns may only play a small role in a country’s decision whether or not to stage the Olympics. The desire to host the Games may be driven by the egos of a country’s leaders or as a demonstration of a country’s political and economic power.

This seems to bear out for megaprojects in general. In Flyvbjerg’s model, decision making is blinded by four “sublimes”, boons of a nearly spiritual nature that distract from more sober cost-benefit analysis. They are:

  • Technological Sublime: “the rapture engineers and technologists get from building large and innovative projects”
  • Political Sublime: “the rapture politicians get from building monuments to themselves and their causes”
  • Economic Sublime: “the delight business people and trade unions get from making lots of money and jobs”
  • Aesthetic Sublime: the pleasure designers “get from building, using, and looking at something very large that is also iconically beautiful”

The Olympics are surrounded in so much mystique and hype that we can easily imagine an aggravating effect. You are not merely building a monument to yourself, you are building it to human glory. The whole world will come to see what you’ve designed. And so forth.

The effect can further be accentuated by a sense of historical importance or sense of inadequacy. As Brazil’s president described his desire to host the Olympics, “We’re not this tiny country people thought.” Or as Sepp Blatter, one of the IOC members said following the president’s speech: “I honestly think it is Brazil’s turn… It is South America’s bid. This is a continent that has never held the Games. It is time to address this imbalance.”


There are many reasons mega projects go wrong… but the astute reader will notice one cause conspicuously missing from this entire discussion.

I do have a follow-up post planned, but in the meantime, it’s left as an exercise to the reader. To win internet points, you’re welcome to email me or share your bad take on Twitter.