Trust Experts vs Trust Knowledge + Reason + Candor

I previously wrote about experts vs non experts on Covid, but I now think that is not quite the right framing. When I’m deciding who to listen to on a topic, I am hoping to find someone with domain expertise (knowledge) AND great general reasoning skills AND who is truthfully sharing their best information. The problem is that on most topics, this person doesn’t exist. Covid is particularly challenging because its such a huge, fast-moving situation that no one can be an expert in all parts of it. Everyone’s domain expertise is partial at best, and there’s no guarantee that the most knowledgable people are also wise and honest. Instead, I must settle for listening to many people with varying levels of knowledge, reasoning, and candor. The tradeoffs among these three virtues are what I was trying to get at with “expert vs non expert”, but reality is messier than that binary.

The responsible-sounding thing to do is to trust the experts at the top of relevant scientific institutions like the CDC. But leaders like that haven’t had a great track record on Covid. 

Now-deleted tweet from 2/29/2020

People get to run institutions because they are skilled at running things or at maneuvering to get and keep top jobs- not because they are the most knowledgeable expert, not because they are the smartest, and definitely not because they tell the whole truth all the time. 

Before Covid, hearing that someone was at an agency like the CDC made me trust them more than someone with similar credentials (say, PhD and publications) in academia or the private sector, because I thought they would be more knowledgeable. Now their job title makes me trust them a bit less than otherwise- I’m no longer sure they are more knowledgeable, but I’m quite sure they are less candid. Former leaders, like ex-FDA head Scott Gottlieb, tend to be better- but they are always angling for the next position, which influences what they say.

One other update from Covid was the importance of learning from data and making accurate forecasts (and I say this as someone who already valued those things a lot). These are their own fields with their own domain experts (statistics / econometrics / ML/ et c), but learning from data is also a powerful general reasoning skill that works across many domains. Those who excel at it have a real shot at beating knowledgeable domain experts at what seems to be their own game- we’ve seen it in baseball, and now in public health (of course the learning can go both ways here, with subject-matter experts picking up data skills).

All this explains why the Rationalists have been ahead of the curve on Covid– great at general reasoning (including forecasting), honest to a fault, and so just need to pick up domain knowledge. Rationalists starting with partial domain expertise, like biologists and doctors, had a big head start.

It also explains why pseudonymous Twitter accounts were sometimes way ahead of the game. From the boring responsible point of view, this sounds crazy- would you really trust a Twitter anon over the Surgeon General? And trusting a typical Twitter anon would be crazy. But suppose the Surgeon General and the CDC Director actually do have anonymous Twitter accounts- wouldn’t these be amazing sources of information? They would have all the knowledge from their day job AND the ability to share their honest assessment without worrying about angering their boss, causing widespread panic, or inducing an N95 shortage. I think some people rely too much on the official/formal/institutional sources because they are choosing “who is higher status” or “who will others not blame me for listening to” rather than “who is most knowledgeable & accurate”. 

The opposite problem is also widespread, choosing first and foremost someone who is a “straight shooter” who “calls it as they see it”- but doesn’t see it well, because they lack domain knowledge and/or reasoning skills. Worse, people often mistake outrageousness for candor and wind up following those who are dishonest as well as ignorant. Disagreeing with the establishment is fine but making a big deal that “I’m telling you what the establishment won’t” tends to be a bad sign.

By ‘candor’ I mean more than not lying- to be a useful source you need to publicly share information, and keep doing it; many experts would never lie but simply don’t have the time or inclination to share much publicly apart from writing the occasional journal article.

To find accurate information, it is not enough to simply “trust the experts”, and worse to simply ignore the experts. Instead we must continually seek out new sources and re-evaluate old ones, finding the domain experts who are relatively open/honest/clear/brave/well-reasoned, and tracking which of the smart/wise/honest non-experts are doing particularly well at learning the subject.

P.S.- Give Helen Chu a medal already. Courage + expertise make a great combination.


  1. Kevin Klingaman says:


    Who are these anonymous Twitter handles? Love to add them if you have a few good ones.

    1. James Bailey says:

      @__ice9 seems like someone in drug development
      @halvorz biology grad student
      You can see who I follow here, though of course not all 700 have been great on Covid

      1. Kevin Klingaman says:

        Thanks – added!

Leave a Comment

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s