About Ergodicity

Anything that Zeynep Tufekci writes is worth reading, and people like Navin Kabra make Twitter a place of learning and knowledge. Therefore this tweet is worth the price of admission twice over.

But it gets better!

Because the replies took me to this excellent essay on ergodicity:

In an ergodic scenario, the average outcome of the group is the same as the average outcome of the individual over time. An example of an ergodic systems would be the outcomes of a coin toss (heads/tails). If 100 people flip a coin once or 1 person flips a coin 100 times, you get the same outcome. (Though the consequences of those outcomes (e.g. win/lose money) are typically not ergodic)!
In a non-ergodic system, the individual, over time, does not get the average outcome of the group.


… And therefore this set of essays and this newsletter.

As the kids say these days: sorted.

Zeynep Tufekci on Metaepistomology

I know, I know.

Here’s what metaepistomology means:

“the theory of theory of knowledge”

And you should now be asking, “what does that mean?”

The latest post on her Substack (god, I can’t afford to subscribe to all the substacks I want to!) is a wonderful essay on how she learnt about the pandemic last year, and how she learnt about how to learn – but I’ll get to that in a bit.

First things first, who is Zeynep Tufekci?

Zeynep Tufekci (Turkish: Zeynep Tüfekçi; [zejˈnep tyˈfektʃi]; ZAY-nep tuu-FEK-chee) is a Turkish sociologist and writer. Her work focuses on the social implications of new technologies, such as artificial intelligence and big data, as well as societal challenges such as the pandemic using complex and systems-based thinking. She has been described as “having a habit on being right on the big things” by The New York Times and as one of the most prominent academic voices on social media by The Chronicle of Higher Education.


I learnt about her for the first time when I cam across a review of her book, Twitter and Tear Gas over on Aadisht’s blog. I haven’t read it yet, but I still remember this from his review, because it resonated a fair bit:

A point this book makes often is that digital tools mean that networked protests are enabled, and that protests can spring up much quicker than they used to. But prior protests used to be much more organised, because the threshold to start a protest used to be so high that it would take a long time and lots of organisation to hit it – and that meant that there would be an organisation capable of pushing for change after the protests. The digitally fuelled protests haven’t quite figured out what change to ask for, and how to push it, yet.


(We also did a podcast about his review – and some other posts from his blog besides.)

But Zeynep’s writing reached another level altogether (both in terms of relevance and in terms of impact) during the pandemic. This, for example in the NYT (note the date!), or this from The Atlantic.

But her latest post, on the 31st of January, is worth pondering at great length. And that’s because while it speaks about the pandemic, and how she learnt about how serious it is going to be, it also contains lessons that are applicable everywhere else in life.

Please – pretty please! – read the whole post, but here are my key takeaways:

China’s attempts at downplaying human-to-human transmission and the WHO’s complicity in it are of course wrong, but this is also a good lesson in understanding why exponentials are worth learning about – if nothing else, at least because manufactured lies cannot stand up to the steep part of an exponential curve. And no matter your opinion about whether or not we underestimated the current pandemic and its impact, you should ask where else this lesson can be applied:

Let’s call this the Principle of “You Can’t Finesse the Steep Part of an Exponential,” after a Dylan H. Morris quote included in a previous article of mine trying to warn about the more transmissible variants.


Second, this sentence:

Let’s call this the “Principle of Always Pay Attention to Costly Action.” 


Principles of economics: incentives matter. Up until the point in time when Wuhan was locked down, China’s incentive was to try and suppress news about the upcoming pandemic. Wuhan being locked down was drastic action, yes, but it was also a signal. And the signal was that from here on in, China’s incentive was to warn the rest of the world about how severe and catastrophic (both in terms of health outcomes as well as economic outcomes) this virus was going to be.

Why did the incentive flip? Because the costs of downplaying the virus (in terms of being blamed for the origin, the suppression and therefore the inevitable spread) now outweighed the benefits.

Put another way, if China (if not through its statements, then through its actions) is signaling that its message has flipped, well, things must be really bad.

When it comes to political leadership, ignore what they say, and study what they do.

Political leadership doesn’t just mean governments. This applies to every single political unit, from the United Nations down until your family. Actions, as they say, speak louder than words.

Outrage and counter-outrage on Twitter is words. Action is action, and a far more reliable signal.

And I learnt from this post about the criterion of embarrassment

The criterion of embarrassment is a type of critical analysis in which an account likely to be embarrassing to its author is presumed to be true as the author would have no reason to invent an account which might embarrass him.


If the guy giving you the bad news is embarrassing himself in the process, then the payoff from making the announcement must be more than the cost of being embarrassed.

If intellectual honesty is at a low premium today in society (and if you ask me, it has always been the case) then a leader being (or allowing others to be) honest isn’t about morality, it is about the cost calculus.

So, the thumb rule: if the leader of any kind of group fesses up, be very worried. Think of it this way: map out, consultant style, two axes about public announcements.

Is the announcement good news or bad news (that is, is the leadership that is making the announcement going to be benefit from it, or be embarrassed by it)?

Second: Is the news real and credible, or is it straight out of Narnia territory? (Detecting this is a skill, and we should all possess it)

That leads us to this chart:

Three things that you need to keep in mind:

  1. The upper left quadrant will rarely be an announcement. That is why one should study what leaders do, not what they say
  2. If what the leadership is doing (or saying) matches up with our assessment of how bad things really are, get really worried, and start preparing accordingly.
  3. The third is the second last sentence in Zeynep’s post: “Everything we needed to know to act was right there in front of us, but it required not just knowledge, but a theory of knowledge to turn it into actionable, timely information.”

And that, my friends, is the point of metaepistomology.