And yes somebody else would have probably had that thought before, but if not, I consider this to be an entry on the global ledger of knowledge that this blog is becoming a part of. At some point, I highlighted the tension between when inference meets retrieval. Just to recap, inference is the process of creating an indirect proof about a state. This can be done through a range of empirical statements/analytical frameworks, and the causal revolution in economics has expanded that way of analysis to economics. In the process of it, we may end up having linearized our thinking due to the nature of the toolkit inducing us to think that way.
Retrieval, on the other hand, is simply asking a question and retrieving the answer to that question. A retrieval process may involve reasoning which aligns inference with retrieval as it makes the latter reverse engineerable. As in, you can reverse engineer how a retrieval query arrived at a certain answer.
But as inference meets retrieval we need to be very careful to ensure that we understand the
Can we rationalize or understand ourselves, or, how deep into ones own programming can one delve to understand ones own reasoning or motivation to study a specific question? Or, put differentially, it may matter a lot where is the question asked coming from. How come
This is imperative as I posit that human intelligence can only perform a Turing test if they pass the Self Turing Test.
An intelligence that cannot interrogate its own reasoning cannot reliably judge the intelligence of another. So the question is, can an intelligence convincingly explain itself to itself?