What Makes a Professional Epistemologist?

Epistemology – the study of knowledge – is still in its infancy, in terms of rigor. Even though the field has been around for thousands of years, it’s only now catching its stride, due to the comparatively recent advent of computing and especially the widespread goal to build artificial general intelligence (AGI).

This development has raised a crucial question for anyone who wants to be productive in this field: what distinguishes a hobby epistemologist from a professional one? More generally, everyone should take epistemology seriously because a happy life absolutely depends on mastery of rationality – and it’s your epistemology that tells you what rationality is and is not. So here’s my non-exhaustive list of prerequisites.

The first prerequisite concerns rational methods of decision-making. One of the core problems of epistemology is that of finding a rational preference for one idea over another. A professional epistemologist must state in clear, simple, and precise terms, how one forms a rational preference for one idea over another, though without reference to any particular idea. Ideally, the statement is one only one or two paragraphs long. (Find my own statement of rationality here.) One can then test the proposed method of rational decision-making against a problem where the correct answer is already known, to see if the method holds up.

One example I’ve been using lately is – believe it or not – Pop Tarts. Why are they called Pop Tarts? You might guess it’s because they pop out of the toaster, or because the filling tastes tart. Try figuring out the answer for yourself without looking it up. How do you form a rational preference for one of your guesses over another? The content of the specific solution you come up with in this case isn’t important – it’s the method you use to decide. Just observe which method you use and try to write it down. At the end, look up the true answer (you can find it on Wikipedia) and see if your method could have gotten you there.

Second, I mentioned above that epistemology is still in its infancy in terms of rigor. A rational decision-making method is a computational task: you can, in principle, write it down, step by step, in the form of an algorithm. I follow physicist David Deutsch in thinking that, if I cannot program a computational task, I have not understood it (The Beginning of Infinity, chapter 7). By extension, if I have not programmed a proposal for a rational decision-making method, then I have not understood that, either. (Find the implementation of my rational decision-making method here.)

Does that mean Karl Popper was not a serious epistemologist? What about Ayn Rand? No, we can give them a pass: as fields develop, requirements change. That’s especially true for philosophers who began their epistemological work before personal computers were widespread – both Popper and Rand began in the early 20th century. But consider Deutsch’s stance on cosmology:

The scope of science keeps growing as we find ways of making testable theories about things where previously we couldn’t. A prime example is cosmology. [I]f you look in an old dictionary, you’ll see cosmology listed as a branch of philosophy. But if you look in a modern dictionary, it’s listed as a branch of physics. That’s an example of the totalitarian character of physics – that it tends to envelop everything else.
— David Deutsch, https://youtu.be/tzWGfi4XhLA?t=316

Cosmology used to be a branch of philosophy. But now it’s a branch of physics, meaning you can’t be a cosmologist without being a physicist. Something similar is happening in epistemology. I’m not trying to gatekeep the field – there’s just no way around it. Questions such as whether animals are sentient, or whether and how the mind could be transferred to a computer, are currently still predominantly philosophical questions. But it’s already far easier for programmers to make progress in these areas, and to see what the full answers might be one day. It’s easier for programmers to recognize how algorithmic animals can be – and how algorithmic even people can be when they’re not being creative.

So I expect aspiring epistemologists to be at least somewhat competent programmers. ‘Vibe coding’ isn’t good enough – you need to actually understand software and write it by hand. It’s common knowledge among programmers that learning to code has helped them think. The specific language isn’t all that important as long as it’s computationally universal – which, as far as I know, all established languages are. Pick an easy one like JavaScript or Python. Solve some of the easier LeetCode problems, maybe some medium ones. That’s a fine start.

In response to my standard of epistemological rigor, and my requirement that serious epistemologists be programmers, those familiar with Popper’s epistemology usually reply that it cannot be turned into an algorithm because it depends on creativity. We don’t yet know how to program creativity – such a program would be an AGI. And even if we did know, creativity cannot be specified in advance like a conventional program. My response has always been that we can simply outsource the creative parts of our decision-making algorithm to the user, through prompts. Just like programmers have always done. So the limitation my critics invoke isn’t as fundamental as they think it is; in the context of rational decision-making, it can be overcome easily.

Third, a serious epistemologist’s texts empower you to think. Sadly, there has been a long tradition of pretend epistemologists – really mystics – whose goal is to attack reason; to convince you that you are fundamentally unable to be fully rational. Don’t listen to them. Serious epistemologists advocate reason and reject mysticism. The point of epistemology is to figure out what rationality is, not to deny it. After reading a professionally written epistemological text, you feel refreshed and optimistic; you know that your values have a chance in this world; that your problems can be solved; and that reason is the only known tool to make sustainable progress and achieve happiness.

Consider an ‘engineer’ who, instead of solving real engineering problems, tried to convince others that we cannot find any solutions, that engineering is hopeless, and that we should give up; that we could never build any skyscrapers or bridges. You’d rightly refuse to call this man an engineer. ‘Enemy of engineering’ would be more like it. The same is true in epistemology: anyone who tries to convince you that your mind is incompetent to think is not an epistemologist. On the contrary, a good epistemologist empowers you to be more independent. In other words, after reading his texts, you depend less on him. You gain clarity and confidence to rely on yourself to judge and make decisions to live your life. But if you find yourself relying more on the epistemologist after reading his work, that’s a bad sign. If people routinely stop making progress after stopping their association or engagement with him, that’s a terrible sign.

Fourth, a professional epistemologist is responsive to criticism. It’s not enough to work on epistemological problems whenever he feels like it – he makes epistemology his business because he always wants to improve in that field. If somebody raises a problem with his epistemology, he works to address it. If he ignores or evades the problem, or puts it off until he ‘feels’ like addressing it, he’s not a professional.

Fifth, serious epistemologists expect and accept consequences for their errors. Unfortunately, academia has a long tradition of ‘philosophers’ and ‘intellectuals’ navel-gazing in ivory towers. Some of them remain deeply mistaken for years or even decades because they operate inside structures that protect them from the consequences of their errors. Paul Ehrlich (whose last name ironically translates to ‘honest’ in German) is a salient example. An honest thinker makes his theories vulnerable to tests and criticism, and sets incentives for others to test and criticize them. To that end, my site veritula.com offers so-called bounties, allowing you to present an idea and pay others to criticize it.

In summary, serious epistemologists meet five criteria:

1. They can state in clear, simple terms, no more than a paragraph or two, how they form a rational preference for one idea over another.
2. They are programmers proficient enough to turn the first requirement into working code.
3. They empower their readers to think independently, and to practice and master rationality.
4. They are responsive to criticism.
5. They expect and accept consequences for their errors, and make their ideas vulnerable to criticism.