Below is the last of the essays from Medium that I need to back up. I have found Medium to be a major disappointment in every respect. My own essays there got zero promotion, as far as I can tell–they would have been as widely read if I had written them as blog posts.

The essays that are promoted to me by Medium (I am not clear on how they mix human curation with an algorithm) in their daily letter are almost invariably insipid. My impression is that 95 percent of the writers on the site cater only to the dogmatic and extreme left.

In the days when the blogosphere was the main form of self-published writing on the Internet, I think that the decentralized editorial curation process worked pretty well. With the demise of the blogosphere, and what are we left with? Twitter? Ugh.

A few months ago, a commenter on this site predicted that the intellectual failure of Medium would be followed by financial failure. That possibility alarmed me because, unlike all of my other essays, my Medium essays were not originally written on my computer. So I have no backups should Medium suddenly shut down. Hence, these backups.

I hope that Medium continues and that my essays stay there, where they are properly formatted. But going forward, I will place my essays elsewhere, primarily on this blog.

Their conversation is available on YouTube. I was particularly struck by how they approach learning from other people. I think of this as social epistemology. My notes here are focused on the lessons about social epistemology that I took away from watching the conversation.
Ordinary epistemology is concerned with the question of what to believe. I think of social epistemology as concerned with the question of who to believe. I have never delved into the academic discipline of social epistemology, but perhaps my concept of it is not too far off.
A few other preliminary thoughts before I turn to their conversation:
Because human beings are social learners, the question of who to believe is arguably the most important question that we face in trying to figure out what to believe.
Part of the social upheaval that we are experiencing in this century is a decline in people’s deference to authority figures in politics, journalism, academia, and professions like law and medicine. Signs of this social upheaval can be found in polls that show declining public trust in these elites, in the persistence of conflicting narratives that go unresolved (think of the narratives around Bret Kavanaugh,Trump/Russia, or Trump/Ukraine), or the opinions that get confidently expressed on this web site (Medium) without going through any sort of peer review or qualification process.
In short, you could say that we are in a crisis of social epistemology. The heuristic of “trust the credentialed expert” no longer works in the age of the Internet. But what does work?
Tyler Cowen and Robin Hanson argue that rational truth-seekers would not disagree in the way as much as we observe people doing. Ironically, I find it possible to disagree with their argument. But I think that their paper is a useful reference point for those of us interested in social epistemology.
Lesson 1: It helps to have customers who know how to buy
At minutes 6–8 in the podcast, Collison explains why as an entrepreneur he started a company that did payment processing. He points out that the reaction to his idea was binary, due to the fact that there were many payment processing systems already in use on the Internet at the time: people without first-hand experience using those systems thought that starting another one was not a promising idea; but entrepreneurs who had tried to incorporate other payment processing systems into their web sites, giving them first-hand experience, were dissatisfied with existing systems, and they gave Collison encouragement.
Collison points to Slack and Zoom as companies that also entered market segments that already were crowded. Yet they were successful because they solved some of the pain problems with existing offerings. And once again, businesses already were using collaboration software and videoconferencing software.
Collison makes the point that when you enter an existing market, you already know that there is demand. More important, your customers already know how to buy what your are selling. This is especially valuable when you are trying to sell to businesses, because getting a business to gear up to make an entirely new purchase category takes tremendous time and effort.
In a generally-forgettable book that I wrote many years ago on entrepreneurship, I said that it is unwise to position your product as totally new and original, particularly to business customers. Instead, I recommended positioning it as “Like ____ but ____” where you fill in the first blank with something that your business customer is already buying, and you fill in the second blank with an attractive feature or price point.
Note that positioning your product relative to competitors does not preclude your seeking to dominate a market. It just takes away the presumption that the path to dominance is to create a new demand from scratch.
If you have something new, it’s easy for potential customers to say that they are “interested.” But it is a better heuristic of social epistemology to put more trust in people who are actually buying a related product. That tells you reliably that they have a need for what you are offering.
At minutes 9–14, Collison describes how it feels on a daily basis to be an entrepreneur. He makes the point that you are just about as stressed whether the business is doing well or doing poorly. In either instance, he points out, you are trying to do as much as you can handle. Even if you have an “Aha!” moment, it often comes in the midst of a typical week of putting out fires, so that its importance only becomes evident in hindsight. I would add that profitability by no means ensures that people with big egos will get along — really successful bands break up.
Lesson 2: Hire people you respect
At minutes 15–17, Collison says that the trajectory of his business was perhaps affected most by high-caliber hires. At minute 18–20, he and Hoffman discuss some heuristics for hiring. These include:
— Is this someone who you respect so much that they could hire you to work for them?
— If you disagreed with this person, would you think it at least as likely that you are wrong as they are wrong?
— Can you learn (a lot) from this person?
Collison describes himself and other entrepreneurs as being disagreeable. This refers to a Big Five personality trait that suggests someone who is willing to go against majority opinion. It describes someone who values honesty more than social pleasantness. I put Nassim Taleb in the disagreeable category (Taleb disagrees — naturally — viewing personality psychology as bogus). Note that on the surface, Taleb and Collison exhibit very different behaviors.
Disagreeables disdain the use of “wisdom of crowds” as a heuristic for social epistemology. Instead, they set very high standards for deciding who to believe. If you are a disagreeable, then before you hire someone it is important to be convinced that you respect them enough that they could change your mind about something.
Also, note that Collison and Hoffman both believe in checking references very early in the process of making a new hire or soliciting a new investor. This is a way of believing who (the reference) as a check on your belief in what (your own impression).
Lesson 3: Don’t fear the weird
At minutes 23–28, Collison discusses that he learned from making the mistake of not buying bitcoin in 2010. He says that his own analysis suggested that bitcoin was a promising idea, but he decided to stay away because it seemed weird. Subsequently, he decided that weirdness is not a good criterion for dismissing an idea.
I would say that Collison made the right mistake. Had he been successful as a Bitcoin speculator, it would have drawn him away from building his business.
Of course, I am a Bitcoin bear, so my opinion is biased. But note that Collison’s company is one of those that backed out of the Facebook Libra project, after initially joining in. My impression is that ordinary businesses (those who do not operate in the black market) are crypto-curious but they have not learned how to buy products in the blockchain space. In some ways, that is the worst possible state for a start-up to confront, because the people you want to sell to will love to waste your time and hate to spend their money. Let someone else suffer through the process of teaching them how to buy products based on blockchain.
At minute 29, Collison discusses the decision to open an office called Remote, meaning an organization that is equivalent to a physical office, but where engineers can work from whatever location they desire. He makes the point that this seems like a logical thing to do, but other tech companies are not doing it. He thinks that perhaps other companies are afraid of the weirdness of the idea. In his social epistemology, if an idea is logical but other people fear its weirdness, that is a sign that it is under-valued.
At minutes 32–37, Collison discusses the diffusion of talent in engineering and entrepreneurship. The San Francisco Bay area is losing its advantage. In part, this is because the challenges of building a normal family life in that area are high, particularly with the cost of housing. But primarily it is because Collison sees other cities and especially other countries (he cites Eastern Europe as an example) as having built up stocks of highly skilled people.
Lesson 4. Give great scientists more freedom
Minutes 41–52 are devoted to the issue of disappointing scientific progress, raised by Collison and Michael Nielsen. Collison says that we are putting 100 times the resources that we put into science in, say, 1950, but getting no apparent increase in the quality of output. It appears that we are misallocating resources in science.
It could be that the process of grant-making in science incorporates bad social epistemology. I hear Collison saying that grant-makers appear to under-rate great scientists making proposals that either seem weird or will take a long time to play out.
Lesson 5. Read what interests you
In the final segment, starting around minute 54, Collison discusses reading. He certainly rejects the heuristic of reading what is popular. He also rejects the heuristic of reading what people close to you are reading, on the grounds that you probably will absorb the ideas by osmosis. Instead, he recommends diving into a subject that interests you much more than it does other people. As an example, he talks about recently reading books that describe the time period when we could blitz-scale engineering projects (a subway system, a skyscraper, a shipyard). He is hoping to figure out why things are different today — instead of getting better at these things, we seem to have gotten considerably worse.
This is another interesting question in social epistemology. I suspect that we no longer have people that you would trust to blitz-scale in construction. The people who had the know-how for that are no longer alive. People who currently have a knack for organization and logistics have found their way to Wal-Mart or Apple or Google, as opposed to construction. I also suspect that at the relevant government agencies we went from a culture of “can-do” to a culture of “no.”