What's on my agenda right now?

You can probably infer a lot about my interests from my writing, but these are my current areas of active interest, for either research, projects, or curiosity.

Epistemic institutions

An “epistemic institution” is one which helps people to find and coordinate around the truth. There are many situations in which true and useful information is harder to find than it should be, and epistemic institutions can help us to solve this. If I want to find out how many electrons a carbon atom has, that information is easy to find. If I want to find out the likely effectiveness of a physical therapy technique to address my back pain, that’s not so easy.

The “facts” model of knowledge works well for the carbon atom, but not so well for the back pain, because there might not be any objective fact about which physical therapy is most effective. It depends on the circumstances of the person, what kind of relief they are seeking, and what kind of effort they’re willing to make. The question asked is “what do we know has worked best for people most similar to me?”, to which there is, or at least could be, a true and useful answer.

It’s not an answer that current institutions are well-configured to find, and this is why new forms of decentralized science will be needed.

Network states

The Network State is a new spin on the idea of starting new nations. It seems reasonable that we might be missing out on innovations in how to organize societies by arbitrarily restricting ourselves to the handful of territorial states that have come into existence via historical processes, and so new experiments could be useful.

History gives us some workable examples—networks of monasteries with their own legal codes, for instance—and modern technology gives us new ways for people to discover common visions, align their interests, and ultimately to build state-level infrastructure for themselves. I’m troubled by what seems like the failure of the modern state to deliver on many of its core competencies, and am interested in how opt-in and highly-aligned societies could simply do a better job.

Collaboration and coordination at scale

At Legra I’m working on collaboration tools to enable highly-scalable networks of action and coordination. We model organisations as intentional graphs: a directed acyclic graph of the things the organisation intends to do, with each edge representing both a dependency between tasks, and reliance by an agent on another agent. This allows for scalability up from small networks of collaboration to whole organisations, with projects simply being sub-graphs.

Human flourishing and developmental psychology

Now is a really interesting and exciting time for cognitive science, for several reasons. The free energy principle and active inference give a workable model for how intelligent agents can model their environments, use attention to tune the precision of those models, and how feedback from the environment can be used to adjust the model weights. Even though this is an abstract model, it seems that evolution ought to approximate it, and increasing empirical work can validate this.

Seeing our world and self-concepts as parts of a generative model gives a number of neat explanations for phenomena that are otherwise hard to account for, and unifies disparate findings from therapeutic practice, spirituality, and psychdelic experience. We should not underrate the possibility that psychology, as a field, could be about to improve dramatically, giving us much better ways of addressing mental health issues and for the cultivation of human flourishing.

These ideas inform my own personal experiments and meditation practice.

The Moonlit Garden is the personal website of Rob Knight.