Adversarial collaboration Art vs science - balancing societal (statistics based) evidence based medicine vs individual practice, decision making in local context Local vs global optimum position Models, frameworks, and protocols of science. Gedankenexperiment - If you had unlimited resources, how would you go about curing one specific type of cancer in one specific type of context, 80%+ success rate. Stage 0: semantic - take care to establish the precise language foundations that can sustainably scale accurate knowledge production from inception to model. Building accurate models that can reliably reproduce physical phenomena, establishing your ground truths a la laws of physics that you can build on, your first principles. Seriously review accepted dogma and evaluate everything in terms of epistemic confidence. Percentages can be handy to change mental models from narrative thinking mode which is much more forgiving to materialist, thinking mode based on real world evidence. Eg - ![[IMG_6597.jpeg]] Medicine challenges - hard to fly real hardware like spacex does when failure is catastrophic- harms real people personally. But cancer such a mortal problem that it’s urgent and the calculus should be ambitious, especially for (otherwise healthy) young people and those that can tolerate side effects well. Those that have a lot of life left to live. Leukaemia a good example of an ambitious protocol that changed the paradigm. Red teaming exercises. Intelligence industry heuristics and models. Test, test, test Constantly revise, update with results and data Most potent real world applicability- Tinkering outperforms design try to make antifragile system design Stage 1: mapping the theoretical ideal combinatorial strategy amongst various overlapping contexts. Contexts brainstorm - Immune metabolic targeted molecular circadian Genetic, epigenetic Modes of investigation- Metabolomics. Investigate cellular models of cancer subtype. Investigate different bio models and their strengths/weaknesses. In vitro, mice, zebrafish, pdx etc. Map out, define, and benchmark metabolic landscape differences to a broad selection of normal cells (especially cell of origin similar).Take care to identify indications of normal cellular processes that would make it difficult to achieve a feasible therapeutic index. Preliminary ranking of therapeutic index potential from known knowledge base. Example - highly glycolytic cancers. Very targetable process but wide applicability to normal cellular processes and potential antagonism to other therapeutic modalities such as immune with highly glycolytic T cells presenting very similarly metabolically. Practicalities - types of medicine Medicines - drugs, natural products, etc Drugs applicable for Bindability Bioavailability TI Enzymatic processes Incentive structures - patentable vs non-patentable etc