The anatomy of knowledge: observation -> induction -> assumption -> logical reasoning -> conclusion -> testing.

Knowledge is inseparable from its production process. Hence the importance of epistemology.

Testing: falsification and risky validation. [https://plato.stanford.edu/entries/popper/]

No knowledge can be proven true absolutely. [https://en.wikipedia.org/wiki/I_know_that_I_know_nothing]

Since no knowledge can be proven true absolutely, it follows that one has to take risk to act upon a piece of knowledge. This is especially true in a market economy. It is possible that repeated actions by many participants can validate said knowledge more to lower the risk. However hidden systematic risk could still increase, since such actions could lead to a self-reenforcing cycle. In order to beat the market, the market participant needs to act differently to differentiate, and he/she also needs to control the risk of being wrong. It is plausible for said market participant to repeat the adversarial process of coming up with bold assumptions and challenging it with existing evidence and reason, until the risk is lowered under an acceptable bound and the reward still profitable.

Human's perception of truth can lead to to a self-reenforcing cycle, which keeps validating the perceived truth until it can no longer do so.

• [https://www.tandfonline.com/doi/full/10.1080/1350178X.2013.859415]
• [https://ocw.tudelft.nl/wp-content/uploads/ElFarolArtur1994.pdf]
• [http://blackout.gmu.edu/archive/pdf/arthur_90.pdf]

If a piece of knowledge concerns an action that is unique to the actions's actor and target, then it usually cannot be gained passively. The actor needs to act in order to examine the pattern of response. It then follows that said actor faces an exploitation vs exploration dilemma: he has to choose between taking new actions to gain new knowledge and taking examined actions to benefit from existing knowledge.

The assumption of a piece of logical reasoning is also a piece of knowledge, which cannot be proven true absolutely. Usually the truthfulness of the assumption is determined by a number of competing factors that tend to change over time.

• [https://en.wikipedia.org/wiki/Lotka%E2%80%93Volterra_equations]
• [https://brilliant.org/wiki/markov-chains/]
• [https://en.wikipedia.org/wiki/Stochastic_programming]
• [https://people.wou.edu/~taylors/g322/baas_2002.pdf]
• [https://pdfs.semanticscholar.org/3fb4/a7631583f9f8d567a6aa6c56a4f33280f1da.pdf]

Over time, the market tends to exhaust the factors contributing to the truthfulness of assumptions that are currently true and amplify the factors contributing to the truthfulness of assumptions that are currently not true, causing paradoxical fluctuation of the truthfulness of assumptions.

• [https://arxiv.org/abs/cond-mat/0402651]
• [https://en.wikipedia.org/wiki/The_Logic_of_Collective_Action]

If we can construct and validate quantitative models that capture the dynamics of different factors contributing to the truthfulness of an assumption, we would have gained a deeper level of understanding of the knowledge that is based upon said assumption. However this is not always feasible. In turn, empirical heuristics are often employed. Human society tends to form fitness selection mechanisms for these heuristics, which could have drastic consequences. It is possible to hack these fitness selection mechanisms to let false assumptions win.

Despite the existence of rational reasoning in favor of being the minority in many game scenarios, herd behavior continues to exist.

• [https://en.wikipedia.org/wiki/Herd_behavior]

Correlation does not imply causation.

• [https://en.wikipedia.org/wiki/Correlation_does_not_imply_causation]
• [http://www.michaelnielsen.org/ddi/if-correlation-doesnt-imply-causation-then-what-does/]
• [http://calteches.library.caltech.edu/51/2/CargoCult.htm]

logical reasoning = proof

proof = program [https://en.wikipedia.org/wiki/Curry%E2%80%93Howard_correspondence]

program = function [https://en.wikipedia.org/wiki/Lambda_calculus]

A computer program is a piece of logical reasoning which requires very little effort to test its assumptions. The basic assumption of every computer program is that the computer which executes it will function as expected, which is the essential point of digital computing.

Interpersonal communication, on the other hand, is a piece of logical reasoning which requires a lot of effort on testing its assumptions, which are other people's subconscious pattern and state, and their assignment of the probability of truthfulness on various pieces of knowledge. Under this view, the words spoken during interpersonal communication is only a small part of the logical reasoning.

• [http://www.columbia.edu/~rmk7/PDF/Comm.pdf]
• [https://www.annualreviews.org/doi/10.1146/annurev.psych.54.101601.145041]

The production of knowledge requires effort in all of its three stages: coming up with an assumption from observations, construction of logical reasoning, validation of conclusion. The distribution of efforts on these three parts may vary, leading to different skill sets in the knowledge production process, which in turn leads to the necessity of teamwork.

If a piece of knowledge is less known, it tends to be less validated. However it could also be more valuable in a market. If a piece of knowledge is well known, it tends to be well validated and thus contains less risk. It is thus plausible for a market participant to make use of less known knowledge in strategic planning and well known knowledge in tactical execution.

• According to George Soros[https://www.tandfonline.com/doi/full/10.1080/1350178X.2013.859415], testing on knowledge involving human interactions is often unreliable because human tends to behave differently when facing a test. Is there a way to solve this problem?
• Can knowledge exist in a distributed manner? Can knowledge be recorded not as logical reasoning but as numerical computations? Can the knowledge production process span across multiple generations? Can one know but doesn't know he knows? Can one know but doesn't know how much he knows? Can one know and thinks he knows but what the he thinks he knows is wrong but he still knows? Can knowledge exist outside one's conscious mind, in the form of subtle preferences and tendencies in the subconscious?

• [https://plato.stanford.edu/entries/epistemic-paradoxes/]
• [https://plato.stanford.edu/entries/plato-theaetetus/]
• [https://plato.stanford.edu/entries/epistemology/]
• [https://ideas.repec.org/a/bus/jphile/v2y2009i2p99-120.html]
• [http://www.scielo.br/scielo.php?script=sci_arttext&pid=S1519-70772007000400008]
• [https://www.hup.harvard.edu/catalog.php?isbn=9780674258969]
• [https://www.emerald.com/insight/content/doi/10.1108/S1529-213420180000023006/full/html]
• [http://philsci-archive.pitt.edu/13108/]
• [https://www.tandfonline.com/doi/abs/10.1080/0267257X.2014.977332]
• [https://www.goodreads.com/book/show/1563563.Understanding_Management_Research]