Three-Toed Sloth

Unsolicited Opinions

Attention conservation notice: Unsupported assertions, accumulated over many years in my drafts folder, last updated in early 2022, now more or less painful in retrospect.

In no particular order, except where they are. I will not elaborate or explain.

• "The Tyranny of Structurelessness" explains a hell of a lot about modern life.
• If the problem is that powerful people are not accountable for their actions, network forms of organization are the very opposite of solutions.
• Increasing returns $\Rightarrow$ monopolistic competition $\Rightarrow$ market failure explains a hell of a lot about modern life.
• Endogamy systematically confounds genetic and cultural inheritance. It would be astonishing if personal-level variables which are entirely the product of social structure, custom, tradition, and strategic action weren't genetically predictable.
• During the 20th century, and in much of the world even today, genetic variation in resistance to lead poisoning during brain development would, psychometrically, look like a heritable general intelligence.
• Multiculturalists who expect different cultural groups to have different values and standards of excellence should not expect those groups to be equally represented in all occupations and professions, especially if people are free to enter and leave different lines of work.
• There is a branch of the wave-function where Margaret Thatcher (trained research chemist and honorary FRS) led a successful push by the international community to address global warming through carbon pricing. (This probably require postponing the collapse of the Soviet bloc by at least a few years, so that her major speech to the UN about the urgency of tackling global warming didn't happen right before the Berlin Wall came down.) In that timeline, pollution charges are regarded as one of the signature policies of neoliberalism.
• Modern civilization may be doomed because (i) we need to act collectively and globally on climate change, (ii) the USA is the indispensable nation at the center of the network, (iii) the Senate and Electoral College give a veto over American policy to a combination of plutocracy, reactionary folk Christianity and paranoia, and (iv) people like me are unwilling to move to Wyoming, or even West Virginia, to counter this. (I realize there is some tension between this and the previous statement.)
• A true Burkean conservative in America should work for the preservation and development of the institutions bequeathed us by our wise and heroic ancestors, serving us for half a century, a century, or more: the welfare state; the regulatory agencies; international organizations and treaties; the military-industrial complex; academia; affirmative action; Hollywood and the rest of corporate popular culture.
• The quantitative social sciences would be in much better shape if the first method everyone learned was $k$-nearest-neighbors, or maybe classification and regression trees, followed by the bootstrap. Linear models and $t$-tests should be, for social scientists, the hyper-mathematical arcana at the back of the textbook which their methods class skipped because there wasn't time.
• "Prophesying upon the eigenvectors" is more transparently about making stuff up than "interpreting the coefficients", and therefore preferable.
• There is something very odd about looking at the distribution of node degrees within a network. It makes the network itself into an odd hybrid of a single interdependent object (a relation) and a whole population. Going "something something ergodicity mumble mumble sampling reveals a process grumble grumble shut up and calculate" makes this only a little less weird to me.
• A good philosophy of probability should explain why all applied Bayesian statistics relies on Monte Carlo, the most frequentist method imaginable. Ideally, it would also explain why Monte Carlo works with deterministic pseudo-random number generators (so the algorithmic information content of the pseudo-random numbers is extremely low).
• I have toyed for years with the phrases "the simulation interpretation of probability" and "the birth of Trumpism out of the spirit of fandom", but I cannot work out what should stand behind them. (I hope they are unrelated.)
• Complexity science is a series of footnotes to Norbert Wiener and Herbert Simon.
• The mythopoetic impulse never goes away, but in populations which have gone through the Flynn/Gellner transition to widespread clerk-ship, it expresses itself not as stories, but as theories which aren't actually explanatory. (It gives us the Oedipus complex, rather than Oedipus.) Every form of progress has its costs.
• Believing that there was a unique and irreversible transition from viewing time as an eternal cycle to viewing time as unidrectional history is taking sides.
• His ideas are flawed by a vicious circle; you are engaged in iterative approximation; I stand firm at a fixed point.
• Dentists are not, in fact, talking their book when they recommend pre-emptive action for wisdom teeth.
• In a well-functioning social institution, incentives, information and (internalized) norms ensure that it doesn't matter who fills various roles in the institution, it behaves the same way regardless. Institutions are always impersonal. It follows that institutions are information-hiding abstractions, if we're thinking like programmers. But if we're thinking like physicists: institutions are collective degrees of freedom.
• Blackouts are humanity confronting its own work, and the effects of its own collective actions, as a hostile and alien force.
• The parable of the Nazi bar is a folk version of the Schelling model.
• No one should be allowed to opine about artificial intelligence unless they've at least spent an hour or two with ELIZA and then stepped through the code.
• "Philip K. Dick and the Fake Humans" doesn't yet explain a hell of a lot about modern life, but it will.