Physicists model heavy-ion collisions at the LHC using fluid dynamic simulations, and to get accurate predictions of final particle correlations, you need to account for the position fluctuations of discrete protons and neutrons within the nucleus.

Not sure what point you are trying to make here. Double ML is a valid approach for debiasing confounding effects.

I disagree. It's vulnerable to all sorts of mishaps. You're now having to worry about data leakage between your treatment group AND your target variable. Casual inference without experiment data is all just a mathematical exercise to make a one size fits all approach to identifying relationships. Yes, correlation has weaknesses. But the name "causal inference" is grossly misleading. It's "well if we assume X, Y, and Z then the effect which we have already assumed is causal is probably around this order of magnitude". And hey, maybe that will help you identify cases where a confounding variable is actually the thing that matters. But you're not going to do better than just doing an analysis on the variables and their interactions. You don't have the brainpower to do this at a scale larger than pretty much all causal methods will begin to fail. It does not offer you the legitimacy the name implies.

I think it confuses far more than it helps.

There are quite a few similarities between the Elo rating system and Kalman filters. I’ve always thought this would be a good way to teach them, because you can start with a simplified univariate case, then modify and generalize from there.

I would love to know more about the similarities. Can you share any resources that utilize Kalman filters in rating systems?

I have a different definition of healthy capitalism, a system that enables people to fulfill their needs/wants while maximizing collective happiness. Wantonly consuming electricity for the purpose of price discovery amidst the back drop of global warming may not be (completely) insane, but it does appear to be a textbook externality, and it’s hardly a slam dunk for capitalism.

Sleep, food, exercise. Those three form a pillar for happiness and success. Some tricks which I’ve found helpful:

Instead of focusing on doing the thing, focus on doing the things that makes doing the thing easy.

For example, don’t focus on going to bed early, focus on avoiding caffeine after noon and winding down your routine an hour before your target bed time. Don’t focus on going to the gym but focus on living somewhere that is close to a gym. Same with eating well, develop a system that makes it convenient to eat well. For example, I do a mix of grocery delivery and meal kits because I know I hate grocery shopping.

By focusing upstream on making downstream execution easy, things fall into place and it all feels more actionable.

My experience in physics was that dissidents were plentiful. It’s easier to poke holes in existing theories than it is to explain why a new theory explains all previous observations. I’m more worried about the reproducibility crisis than I am about scientific group think. Most scientists I knew, including those I disagreed with, were open minded people.

May I ask where you are working now?

As an independent consultant; thinking about starting my own AI company as I can build full MVPs with ChatGPT and vector DBs in like 2 weeks and there is so much demand.

> I can build full MVPs with ChatGPT and vector DBs in like 2 weeks

To do what? What's the value proposition with AI these days? I have no clue because all I see is the hype.

Would you consider working with others? I'm curious about this whole product area

Maybe but right now I have some prototype in the works and a potential customer who requested it so any colab would have to wait until that is done.

It’s pretty misleading label for the estimate. It’s the age you get when you reverse extrapolate current physics to the point where current bodies converge in space and time.

It’s possible space and time continue beyond that extrapolation point. No one knows.

If there is no way to know what is earlier than that point, then it meets the definition of the starting point for the observable universe.

Similar to what you see today, but wide spread acceptance and adoption. Companies that that are worried about privacy issues today will get over their reservation by 2025. Many software jobs will provide some kind of access to one of big tech models.

I think terms will emerge around querying LLMs similar to how Googling became a term. Peoples’ behaviors will change and they’ll query models more regularly for more use cases. AI generated contributions will become more culturally acceptable.