My talk at MIT Sloan School on simulations and science

Here is the video, a bit short of an hour, you may recall this was my abstract:

What happens when a simulated system becomes more real than the system itself?  Will the internet become “more real” than the world of ideas it is mirroring? Do we academics live in a simulacra?  If the “alt right” exists mainly on the internet, does that make it more or less powerful?  Do all innovations improve system quality, and if so why is a lot of food worse than before and home design was better in 1910-1930?  How does the world of ideas fit into this picture?

By the end I considered whether we might make science better by first making it “worse.”  I also covered The Phantom Tyler Cowen, and whether attempted refutation is the best way to approach a new idea.


The Kuznets Curve! TC realizes that to make something better we must first make it worse, due to the laws of entropy...a deep thought that will go over the flat top of most of you.

Yes, this thought did indeed go over my flat top. What did you have in mind Ray?

@Asher - asking the right questions is a sign of high IQ, Asher ;-) What I have in mind is this: to order something into a 'better' state requires energy. This means by the laws of entropy you will actually create more disorder somewhere else (Google this, it's not just a good idea, it's the law). So overall the universe gets more messy (simple analogy, mining energy means roughly 67% of energy is wasted in the process, but you are left with 33% of good, concentrated stuff). Likewise, the reason your desk is messy is due to entropy (and fixing it takes work), likewise, 'improving' the environment via dam building and the like will at first destroy it, until you make enough money to repair it again (Kuznets Curve) and are left with a 'restored' environment plus a dam and other 'good improvements'. Strip mining coal and fixing the scars is another example. The analogy extends far and wide, since it's based on physics (2nd law of thermodynamics). It holds for social organization as well: to flesh out a good idea requires a lot of work and discarding bad ideas, bad organization. For this reason, btw, some social scientists say that Big Government is actually a more efficient form of political organization than 'anarchy' since history has shown it to always arise out of more informal organizations (history being the Great Filter of good/bad ideas). Thanks for asking!

Sorry Ray but the second law of thermodynamics doesn't apply to societal changes - we don't need a revolution every time we want to make things better. And I don't even know if it is true that industrialisation follows the same process, sure the early days of the IR were grim but so was living off the land as a poor agricultural worker in those days. Maybe working in a smoky industrial factory was actually an improvement for everyone concerned.

And of course, the 2nd law doesn't even always apply in physics - the big bang is an example of that of course.

I think there is some truth to Ray's point here. In human affairs, as in physics, disorder and chaos are the default, and it takes effort to organize things otherwise. He does call it an "analogy", and it's an apt one.

As for this:

"of course, the 2nd law doesn't even always apply in physics - the big bang is an example of that of course."

Isn't that sort of a problem for physics that hasn't really been solved?

Good rebuttal by dan1111. Indeed a closed living system like a living cell is a 'proof' that the 2nd law does not apply, if you define the boundaries for the second law as the living cell only, and only as long as the cell is alive. Another example is the 'sun shining' while the sun is not yet imploded. My analogy btw is not original. If you have doubts about it, take it up with Freeman Dyson (or his living equivalents like Tyson deG). Messy, unfragmented HD = 2nd law in action, same with unkempt bed.

Bonus trivia: by God, he's still alive!

Wow, this sounds like a complete abuse of the concept of entropy. Typical of social scientists glomming onto a principle of physics and applying it willy-nilly to enhance their pet beliefs while sounding more sciencey.

Entropy applies to closed systems. So long as we are importing energy to make our systems work, entropy need not apply.

In fact, the entire story of life is the story of rapidly increasing complexity - order out of disorder. Yes, in the grand scheme of things that order has to come from increasing disorder overall, (and in fact, increasing local order accelerates the overall rise in entropy), but any local system can have increasing order or complexity without first having to have more local disorder - the disorder can be exported as high entropy waste. A solar panel takes in high entropy energy, a black box converts it to work, and the waste heat radiates away. The overall result in an increase in entropy, but to whatever is inside the box, the story may be a constant improvement with no need for reverting back to disorder.

Even your desk example may be wrong. There is plenty of evidence that messy desks are actually an example of the growth of information and complexity. If the mess on the desk were truly random, you would have a point. But there is lots of evidence that messy desks are actually a form of organization. A blank desk carries very little information. a messy desk carries a lot, and for some people may be a critical organizational scheme. Have a look at the desks of some of the most successful people in modern history. Creative types in particular may utilize their desks as a form of external filing system.

The lesson of top-down government is likely the opposite of what you suggest. Top-down government destroys information and complexity. A price contol's worst feature is that it destroys all the information about supply and demand outside of the mandated price window. The organizational structure of government, is nowhere near as complex as the natural organizations of people left alone. The creativity in government projects is nowhere near the level of creativity generated by millions of human brains left to achieve their own ends. And so it goes.

Dan is a sharp and perceptive guy.

I therefore expect him to get bored and leave soon.

(unless he comes to enjoy mocking people.)


Guilty AS Charged.

"The Kuznets Curve! TC realizes that to make something better we must first make it worse, due to the laws of entropy...a deep thought that will go over the flat top of most of you."

Where to start?

Some problems with theories and curves: one, assumptions, estimates, forecasts, etc. may not perfectly mirror the actions of millions of (rational, typically motivated, etc.) actors/players; may not have factored everything; may not have identified interventions/interferences; theory isn't reality; etc.

Who is this "we?"

Of course, the "we" that have been running things since say 1913 whether wittingly or unwittingly have generally destroyed thinfgs ti save them.

Were the "laws of entropy" passed by Congress, referendum, or imposed by activist judges appointed by geniuses like Monica Lewinsky's boy toy Clinton, Obama, Bush?
approach the real.

Finally, top-down planning doesn't work. And not simply because our mandarins ain't as smart as they think.

Confession: last Physics course I took was over 50 years ago in high school. Good thing they graded on the curve.

or mebbe not to late for the sociology dept
to boldlyreenvision their priors
the middle class understands that when trump pretends to wipe
snot off the Canadian fellas jacket he is intentionally creating
the meme of the day by manipulating the media
when will the media understand they are being manipulated?
its a noisy sandwich narrative that gets in the way of
what they used to call the news
the sandwich narrative is death
death to the sandwich narrative

that will be 189 dollars

so its sorta a feedback cycle that works mostly in trumps favor
like a trump casino
perhaps someone should tell the media
the media thinks trump is dumb
but trump thinks the media is dumb
but if trump consistently successfully manipulates
the media doesn't that mean he might be smarter than most
of the media
that will be 5$
or we cant be responsible for what happens
to your files

"Do we academics live in a simulacra?"

Jean Baudrillard called from the 1980s and wants his particular brand of postmodernist balderdash back.

Keeping in mind Wikipedia always seems to sanitize ideas I've found, reading doesn't seem to be balderdash, it's actually pretty obvious and conventional. For example it's a well known truism that people's conception of reality is based on reconstructions by the brain (since reaction times are not fast enough to cope with actual events). But I trust, he being a modern French philosopher, he wrote a lot of awful and metaphysical prose I'm sure.

The ideas are interesting, the prose is bad, but luckily S&S is v short

Isn't the singular "simulacrum"?

"why is a lot of food worse than before and home design was better in 1910-1930?" - I disagree, both the mean and variance are much higher in both categories under almost any definition of quality you care to create. I think what is going on here is false nostalgia - Tyler is comparing the highs of previous times with the lows of today.

As noted by Prof. Cowen, this abstract has already been posted - and discussed - before at MR (81 comments), a number of which address your point.

Isn't simulacra the plural of simulacrum?

We offer:
Loans for trucks, trailers, tractors, boats, car purchase and personal reason.
Business startup and other types of loans are available.

We are dedicated in providing alternative funding options to those that don't fit into the conventional lending system.

If you are interested in getting a loan text me or contact me so we can proceed with getting you approved and funded.

"to make something better we must first make it worse, due to the laws of entropy...a deep thought..."

too deep for me, as being a universal truth.

If one simply STOPS doing harmful/negative actions -- that can indeed solve some specific problem at hand with no additional energy required (common metaphor: if your head hurts from constantly banging it against the wall -- simply stop doing that)

I suggest that the authors read The Jungle by Upton Sinclair. A book written to inspire workers to unite in a socialist utopia became, to the general public, a need for food safety and gave birth to the FDA. Socialism had limited appeal to the public, but food safety hit them in the stomach.

Next look at the collectivist farms of Stalin and Mao or the oil wells in Venezuela. The collapse of these efforts lead to black markets and corruption. Authoritarian governments then crack down with laws against the enemies of the people. But government agents are often the most "corrupted" by the black markets. Eventually, even authoritarian governments recognize the failure of such policies and seek to create more efficient markets will still maintaining control through regulations or laws.

Mafia Dons, Russian Oligarchs, Chicago Politicians and the Chinese Chairman of Everything, share some traits. They gain in power, prestige and wealth as the empires they control grow. However, prosperity under such regimes is a double-edged sword. Per Gary Becker, as incomes rise the people want to acquire more goods. In some cases, these goods are material things, but they also seek cleaner air and water, and in some cases, increased freedom. Laws or rules are then made that can maintain both growth and control.

Rent seeking is a vital part of the game, but how to collect and distribute those rents while maintaining control is a delicate balance.

For a clear example look at Chicago real estate taxes. The government passes high taxes, and then key politicians offer their private legal services to help some appeal and lower their tax bill. The disparity between the laws and the enforcement of those laws is vast. How is the system maintained? The economic rents that these politicians gain are spread through the system. Contributions and support are given to other politicians. Jobs and favors are granted to voters. Various interest groups are supported in return for turning a blind eye. Various "watchdogs" have little incentive to enter the potential fight. Too many birds are dipping their beak, and they are afraid to poison the watering hole.

China appears to be about rent seeking. Liberalization might increase the growth of the economy but why risk losing control of the watering hole. Too many people benefit from the current system. You might occasionally need to distribute some of the gains from the system to key players, but bureaucrats rarely want anything but incremental change. Why? Because the risk of rapid change can often have unintended consequences.

To maintain and distribute power, the State has an interest in gridlock. As long as people understand how the game is played, and they are willing to play the game, the game continues. Gridlock is part of the game. The government having the levers to create and release the gridlock creates rents that they can distribute and maintains their power.

Then look into Public Choice Theory about how the birth of these laws work in practice. The passage of laws tells you little about how the implementation of the law will work. The rhetoric and laws of Stalin, Mao, or any government does not tell you about the rise of black markets, political corruption, and a host of unintended consequences.

Sorry posted in wrong place

You're in your 50's? Have you recently suffered a head injury? I suggest - in all seriousness - that you consult a medical professional. There really seems to be something wrong with your cognition. First, simulacra is the plural of simulacrum; we do not say "a simulacra". Second, no one beyond age 15-16 should confuse the actual physical world with ANY abstraction. And yet you make absurd statements like the alt-right *existing* on "the internet", that the internet (an abstraction) will become "more real" than another even more vague abstraction, that people live in abstractions, and that the proposition "All innovations improve system quality." (or any similar unqualified generalization) is worthy of discussion. Please disregard the above if you have been already diagnosed and are attempting to deal with your neurological issues. If so, my sympathies, but perhaps an hiatus is appropriate.

Li, we REALLY need you to stick around and post more. Please.

1) ChrisA is correct.

2) Cowen uses the examples of, twitter and Linked-in as services that first got worse before the got better and reads of examples by those who thought the quality was bad at first but each of those were new so those examples aren't applicable. The same with the three minute song.

3) Cowen doesn't mention prices just as he didn't when he used flying as an example of how it has beccome worse since the 1970s. Of courses prices have dropped sharply and the number of deaths per mile has plummeted. (A student asked about price and then Cowen responded.)

"why is a lot of food worse than before and home design was better in 1910-1930"

no it isn't. The basic function of food is to provide calories, and food is cheaper and more readily available than ever before. The basic function of housing is to provide shelter from weather.

Now, some aspects of housing and food are more expensive now than in the 1920s because of scarcity of the inputs. You'll find pine hardwood floors for example in many old houses. You certainly can get pine floors, but it will cost you. Are pine floors really "better" than synthetic alternatives? This is a matter of taste. Also, I would not call plaster and lathe walling "better' than modern drywall. I would not call 1920s era knob and tube wiring "safer" or "better' than modern wiring! Heck no, that knob and tube wiring causes fires!

Perhaps its not true that all innovations improve system quality, but that is chiefly because those innovations exhibit characteristics like networking where people adopt a technology because other people are using that technology (think Beta/VHS, Windows, or Facebook).

Products that are simply luxury items, like the ability to whine to your friends on Facebook, are simply that: luxury items people indulge because they have too much time and money. People have too much time and money, chiefly because basic food and shelter is cheaper than it ever was.

“Better” by what measure? For what purpose? For whom? Etc.
Standing unsupported by a system of values, for a specific individual in their specific context, ‘better’ is not just meaningless, it is the death of meaning.
It purports to say much, while saying nothing at all, yet how seductive a word it is.

For starters, I'd say the massive reduction in deaths from food poisoning and stomach cancer is 'better'. Also, the widespread availability of fruit from around the world is better. Also better: foods that can be prepared quickly. For those of us unfortunates who do not have live-in cooks and who work for a living, the ability to prepare a nutritious meal in half an hour is certainly 'better'.

I recently saw a video of Matt Yglasis talk to Tyler about The Great Stagnation when it came out in 2009. Tyler begins that unlike his grandmother the only innovation he has seen was the internet. The starting point for seeing new things would begin for Tyler to be when he was 4 years old in 1967. He then saw the microwave oven in the 1970s along with cable televison and the home computer and video games that decade.

Then the walkman, VCRs, CD players, the cell phone, email, blogs, wikipedia, Google searches, machine translation, i-pods, widescreen televisons, and smartphones. I coun't 15 not 1 innovation and that includes the all mighty microwave.

Oh! I almost forgot about junk bonds! So make that 16.

Then there is this later in the 2009 interview:

Cowen: " If you look at total factor productivity, that's the amount of increase in living standards is coming from new ideas, and that is sharply DOWN from 1973 - its almost zero."

The 1990s weren't special as Tyler claimed. The rate of the increase in TFP was the same from 1982 through 2005. 32 years.

So I had to fact check since obviously wrong. FRED is here to help. With an index of 2011 = 1, TFP (Total Factor Productivity) was:

1973 0.7
1980 0.7
1990 0.8
2000 0.9
2009 1.0

There was a major increase increase in TFP and was zero only from 1973 to 1983.

I like Tyler's idea of "competing simulations"?

Inspired by Harold Demsetz, he also poses a good question: should simulations be "open-source" or "owned"?

The 'more real'simulation may discover that Alt-L and Alt-R were figments; artificial groupings of various sub groups looking for yet another government check. All parties thus indistinguishable as welfare bums.

"why is a lot of food worse than before and home design was better in 1910-1930"

We stopped by the Korean Market yesterday, and having time and being hungry, got lunch at the restaurant in the corner. Apparently the lady didn't really speak English, and I didn't get number 7 (spicy beef soup) but we got number 3 (hot bi-bim-bap). That's another good choice. Karma. It came in a smoking hot stone bowl. With six kinds of side-dishes. The best food I've had in a while. At $10 per person also very high ROI.

And man, that stone bowl. It was still too hot to eat at the last bites. Koreans. And why don't they have this in ski towns?

Maybe someone was eating this well in 1910, but it wasn't as easy in America as it is now.

The double the scientists thing isn't scary, especially over 13 years, especially as billions of humans are exposed to higher education for the first time.

Worry about it when the world is at 50% college completion.

On making things worst to make them better, I would prefer a different way to say it.

If you are pretty sure you are far from the global optimum it is important to add randomness, to avoid traps at local maxima.

But adding randomness is not actually "making things worse," it part of a change and measure loop,

I was at this talk and I thought the critical comments were odd and felt like academics needing to prove his comment on nitpicking correct. Maybe my understanding is not 100% but the idea of "worse to get better" seems very analogous to the disruption theory of Clay Christensen (who teaches down the block at another business school) and a lot of the misunderstanding seemed to come from soem kind of assumption that we were talking about one dimensional value judgements, and not a more vague notion of utility that is based upon the multiple dimensions that a good or service can be valued, which are subject to tradeoffs.

Then again I dont go to school on Mass ave, so what do I know....

Homes, air travel, food... These have one thing in common. They are all things that are VASTLY better today than in 1910 for the middle and lower classes, While perhaps not being quite as nice for the wealthy.

It may be true that giant mansions in the past were aesthetically more pleasing than giant mansions today - but average homes for average people are incomparably better today than they were in the past. The average person has more than twice the living space, air conditioning, central heat, better insulation and fewer drafts, better, cleaner flooring, more windows, etc.

In 1950 the average home was a salt-box bungalow of about 900 square feet, partitioned into three bedrooms, a living room, a kitchen and a single bathroom. The flooring was usually linoleum with area rugs thrown over it.

Today the average new home is closer to 2,000 square feet, with unique architectural features, multiple bathrooms, wall to wall carpeting or hardwood flooring, master bedroom with ensuite, etc.

It's true that flying transatlantic on Pan-Am in the 60's was glamorous and you were treated to 5-star service. But only the rich could fly. Ask the working family in 1960 if they like air travel, and the answer would most likely be, 'I don't know - we've never been able to afford it.'

I don't know what food was like if you were in a luxury restaurant in 1910, but I know what it was like if you were middle class - you ate a lot of canned goods (canned as in, 'canned at home from the garden'), or you bought food from the local grocer and hoped you didn't get food poisoning or worms or other maladies. Forget frozen foods, treats, or fresh fruit unless it was local fruit in season. You ate a lot of poor cuts of meat, bread, and local vegetables.

In short, this sounds like a list conjured up by a rich person lamenting the loss of the days before the hoi polloi ruined everything. I'm not trying to disparage Tyler, But to middle class me, the idea that any of these things have gotten worse over time is absolutely baffling.

My Phantom Tyler Cowan wouldn't ask a bunch of silly pop-sci questions like those, but then that Phantom was formed years ago when the thing it was modeling wasn't trying so hard to be a pop-intel thought leader.

Comments for this post are closed