Thought
<!-- gdoc-inlined -->
An intellectual life driven by the feeling of insight is actually deeply motivated by emotion. Ideas have emotional valence, and survive on that basis.
The frame from memetics is that the ideas that survive are triggering emotions which lead people to remember and share them. Rather than a being a world with deep feedback loops with reality, ideas and interestingness tend to exist in a social reality shaped by emotional forces.
Particularly triggering ideas - ex., category theory, information theory, and memetics, have emotional force that comes from a few sources.
One is their generality - the promise is that once understood, the breadth of the implications will be stunning. They’ll speckle your worldview and your understanding of reality with an array of changes, mainly by being about incredibly deep and fundamental processes. A second is their ungraspability. Generating awe is much easier when the size and scope of an idea or space can’t be fully grasped, but when the partial grasp feels transformative.
The depth of the update is often a function of depth of connections in the mind experiencing the insight. If you’ve connected the subjet-object distinction to the embedded agency problem in reinforcement learning, both of which are connected to simulationism and the way that the simulated to act to escape into their reality, you have a very different emotional experience of everything you learn in those domains. There are suddenly implications of your learning for more parts of realty than you can grasp in working memory. The experience of that set of potential insights is shattering to somebody who has already set up those connections.
But the experience of somebody who has not created this web of connections is that a new insight stops in the domain where the learn. They hear about power law distributions and isolate the insight to working on probability theory and statistics problem sets. They haven’t created a dense web of connections which would back the emotion of insight in the face of the ideas they see.
Optimization pressure also disincentivises the kind of creative process that comes with asking what the full ramifications of an insight are. On reason people don’t feel abstract insights particularly deeply (or don’t abstract to begin with) is that they can’t concretize it.
The self-unaware promotion of content because of its intellectual memetic value should motivate an examination of our intellectual senses. We should ask why we’re so titillated by the insight we’re seeing. What is means for something to be deep is usual for it to be general and very impactful. Some topics are deeper than others.
Do I spend time on information / knowledge / insights that are likely to remain true over time? Are they about deep and immortal topics like the nature of reality or the nature of information, or are they new that’s local to a specific context which can be safely abstracted over or be ignored by any agent that only cares about general knowledge?
Source: Original Google Doc