Note: These are incomplete, unorganized ramblings. But, I must go eat lunch, so I am publishing anyway should I never return to complete. A conflict of interest typically refers to when a person with power within a system works on behalf of that system on a matter which that person…

The U.S.’s recent negative shift in public attitude toward foreigners has hurt its human capital: Many U.S. companies are reporting new, terrible difficulties recruiting. This yields wage inflation, which merely burns money because there is no associated incremental productivity change. Competitive businesses are being created in increasing velocity outside the U.S.; e.g., Canada, China, and India. This is bad because the U.S. is not capturing the value of the companies that could theoretically be built in the U.S.

Musicians should distribute their music everywhere for free to maximize brand affinity. Their audience is more monetizable (e.g., paid product promotions on social media) than their music ever was. The business is celebrity, not music, and Kim Kardashian (est. $250M net worth) competes with Taylor Swift for your attention. Music can build an audience just like parasocial interaction on social media. Restricting any distribution (by charging for content) dilutes that brand value. If musicians choose only to compose and perform and eschew brand development, then they can not complain about the music industry because music can earn more than ever before.

Note: These are incomplete, unorganized ramblings. But, I must stop to eat lunch, so I am publishing anyway should I never return to complete. I disagree with Chomsky’s universal grammar (UG) theory, which states because every language has a recursive grammar, then human brains must have an inborn cognitive disposition…

Note: These are incomplete, unorganized ramblings. But, I must go eat lunch, so I am publishing anyway should I never return to complete. Writing and, to a lesser extent, reading trains our mind’s languages models (i.e., statistical word retrieval based on previous words) to articulate faster. Trains via language models…

Note: Please pardon any incompleteness, for I wrote the following conjecture from my phone in a parking lot. Idea for performing NLU with neural networks: An LSTM network that generates a parse forest from text input. Then, separately perform A* search on the network output to construct the k-best parse trees with concurrent semantic reduction of semantic…