Posts

Comments

Comment by Max_Daniel on [Link] Beyond the hill: thoughts on ontologies for thinking, essay-completeness and forecasting · 2020-02-14T19:19:08.573Z · LW · GW

[Thoughts on what to do if there is an ontological mismatch between one's thinking and a tool]

  • When I saw Jacob present a version of the OP in person, the discussion focused on cases where the correct response is to use a different tool, ideally one that matches the natural ontology of ones thinking. E.g. when using a whiteboard rather than a Google doc to express thoughts most naturally expressed as a mind map.
  • But I think it's important that there are other cases where it can actually beneficial to 'learn how to think in a different ontology'. I think this is quite common in pure maths, but also shows up in more everyday situations: e.g. initially I found it quite counterintuitive to use, say, Emacs org mode or LaTeX, but after I had payed the fixed cost of adapting to the ontologies imposed by them I actually think that it made me more efficient at some tasks.
  • Similarly, I think it's useful to be able to translate between different ontologies. To learn this, it can be useful to deliberately expose oneself to ontologies that seem unnatural/bad/cumbersome initially.
Comment by Max_Daniel on [Link] Beyond the hill: thoughts on ontologies for thinking, essay-completeness and forecasting · 2020-02-14T19:18:07.993Z · LW · GW

[Thoughts on the term "ontology".]

  • In philosophy, ontology refers to the subfield aiming to answer the question "Which things exist?".
  • Perhaps as a consequence of this, hearing "ontology" makes me think of questions like: What are the primitives or building blocks here? E.g. in a spreadsheet, there would be cells; in a graph there would be vertices and edges etc.
  • But I think the important things Jacob is talking about show up as answer to the different question of which relations there are between these primitives. E.g. the important thing about a spreadsheet is that each cell is unique identified by a pair of numbers, that you can perform computations/functions on the values of cells; the important thing about a (certain type of) graph is that every edge has exactly one vertex as source and exactly one vertex as target.
  • Brainstorm for alternative terms: structure, conceptual structure, conceptual apparatus, conceptual scheme, conceptual toolkit, relations, relational constraints, conceptual landscape, ...
  • (The math version of my complaint is that "ontology" makes me think of set theory, whereas I think the important bits are more naturally visible from a category theory point of view.)
Comment by Max_Daniel on S-risks: Why they are the worst existential risks, and how to prevent them · 2017-06-21T10:26:09.016Z · LW · GW

Thank you for your feedback. I've added a paragraph at the top of the post that includes the definition of s-risk and refers readers already familiar with the concept to another article.