Chat reminded me of the costs of constant iconoclasm in academia: my summary Ronald Loui Ronald Loui r.p.loui@gmail.com September 10, 2026 I once said "the bigger the church, the bigger the fire," metaphoricially of course. I meant entrenched, canonical schools of thought that dominate through evangelism and conversion. Please don't burn things down because you can't think of better solutions. But as metaphor for academia, yes. It reminded me that this was not a strategy to be popular. Few ideas were actually met with open arms; most were resisted, or worse ignored. But Harvard taught me to write for future minds, not current ones. So i never cared. Citation counts are great but they are a trap for those who perish in the next paradigm shift. In chronological order, here's the accounting: high school: went after list-based spell checking. suggested human proofreading would be replaced by a lesser spirit. got some CACM attention. college: went after Dijkstra shortest paths. suggested putting stochastics and multidimensionality in the models. won an ACM award. grad school: joined advisor Kyburg going after point-probabiities, Hurwicz decision rule, Bayesian view of epistemology, causality in science, small-worlds MEU and theory-formation in decision theory. Judea Pearl liked some of it, somewhat importantly. grad school and postdoc: went after deductive logic and nonmon in AI. chat seems to think this is the main work even if it is derivative of Pollock and Nute, Hart and Rescher. AI and Law people later loved it, and it's a complete success for modeling the interplay of meaning and reasoning (why Wittgenstein is more relevant than Turing to chat/LLMs). WashU years: went after mathematical semantics, outcome-distribution-defined fairness, non-process social science theories, non-rationale conceptions of rules, game theory for bargaining, and naive economic conceptions of value. AI multi-agent people won awards citing our shallowest ideas here. the Maslow change of calculi, the barista/Gates quip on transactional value (earlier a point about massage, but we can keep it G-rated), the nonstandard utility for process with an investment in non-manipulation: these should eventually shine, though econ is more entrenched in bad ideas than ever. Thaler should have been an ally, and Tversky too, but they were already too famous to care. also went after pub counts in favor of cite counts (but now say citations have become less than they seemed). Cycorp and UIS years: went after Java for cs1, medical records with fixed ontologies, risk analysis that avoided probabilities, and ml predictors ("indecent induction"). some cyberwar ideas (some said cyberwar was not a form of war; i disagree!). got a lot of attention advancing scripting languages, some attention for cyberwar, and no pickup elsewhere. happy to be an early proponent of python for cs1, and prevailing on Cyber Pearl Harbor, but the EMR and milestone-path-probabilities would have made the world better. CWRU: going after predictive AI with the Colvin Diagram as visual auditor (and data set empirical truth finder), going after summary stats generally in favor of plots (but hating on linear regression), going after narrow, non-conversational causal explanation. after some assumptions in OS like smp, cfs, and mt. completely rewrote the Genesis timeline of the Torah to match linguistics, geopolitical history, biology, anthropology, geology, and theology, and other documentary sources. clearly the best work yet, and more of the biggest churches. We shall see how hot they burn. Actually, i prefer churches be repurposed as college dining halls, so perhaps the rebellious youth's metaphor was not really apt. The thing is, the new GenZ really hungers for new guidance on these age old problems. They might have more time for technical precision and philosophical savvy, and a nose for pretenders, proselytizers, rectors, and vicars. So if you are just repeating liturgy, see you in the after-life. I might yet be the one with 72 beautiful minds wanting to hear what i said.