cultus

Also see Oleg's "Effects without Monads: Non-determinism -- Back to the Meta-Language"

https://arxiv.org/abs/1905.06544

It's just a perspective, but Bayesian inference makes the prior explicit, wheareas it is implicit with frequentist inference. You can't not have a prior.

My point is you that you cannot ever actually claim to ground a probability value in objective features of the world. Many events aren't repeated, for example.

Every frequentist technique has a Bayesian interpretation and vice-versa. Confidence intervals are equivalent to credible intervals with certain priors.

There's no definition of probability that doesn't involve philosophy or metaphysics [0]. Calling frequentist stats "objective" really bugs me. There is no such thing. Every inference procedure involves subjective choices.

Of course, frequentist and Bayesian stats are completely mathematically equivalent. The choice just affects our mental patterns.

[0] https://en.wikipedia.org/wiki/Probability_interpretations#Ph...

The geologic evidence for Snowball Earth (multiple ones, actually) is extremely strong. It is almost impossible to explain the re-emergence of banded iron formations without it.

The hard part is explain is how photosynthetic life survived. The Slushball Earth hypothesis holds that there was a band of thin ice near the equator. This has some severe problems though. For one, it isn't climatologically stable. The oceans in SE were covered by "sea glaciers," which were floating glaciers thicker near the poles, which flowed to the equator. These flowed due to the weak hydrologic cycle, with ice experience net sublimation near the equator. These would have overridden any areas of thin or nonexistent ice.

A hypothesis that is getting more traction is that narrow equatorial rift seas (like the Red Sea) would have prevented inflow of the sea glaciers enough for thin-ice refuges to exist. These kind of seas probably existed, because this was occurring as the supercontinent Rodinia was beginning to break up.

A person I know went to Mexico for a semi-elective jaw surgery. It was a great experience: more personalized attention than the US, high quality, and they paid much less.

Yes, you are still legally allowed to pay many thousands of dollars for healthcare if you are laid off.

Cataract surgery is not urgent. In Canada, they ration by need, not the ability to pay. It's also a problem that is not generally characteristic of health care systems. Overall health outcomes are better by the vast majority of metrics.

Even if you have good insurance, wait times can be bad. For a point of reference, someone close to me had a potentially fatal condition and wasn't able to get a specialist appointment for 4 months. That sort of thing isn't uncommon.

Parents are mortals, and modern apps are specifically designed to be addictive. It's the parent's fault, sure, but it is also the fault of the unethical practices of social media. Also, when you're talking about older children and teens, there's really a limit to how much you can actually control them.

People in the healthcare system are doing the best they can, putting their lives at risk to help people, with insufficient resources.

Every should be very, very angry. It should be directed at politicians and corporations, not middle or working-class individuals trying their best.

Not much I don't think. This has a few examples:

https://github.com/rossberg/1ml

There isn't special module syntax. Modules are more-or-less just structs, and the structs can contain types.

Module functors and first-class modules are probably the big feature. Those are more akin to allowing functions between classes, rather than just objects.

The shakeup COVID has caused hopefully will result in some reforms in higher education, which is just such a bloated top-heavy mess with self-serving administrators and boards that it makes the military look efficient.

Great read. It's too bad modules are pretty much an afterthought in most langs.

I really like research language 1ML's approach to modules[0]. This allows monomorphic types to be treated like values, avoiding all of OCaml's module syntax (which can be a bit complex and verbose).

https://people.mpi-sws.org/~rossberg/1ml/1ml-extended.pdf

Bell bottoms still look stupid, though.

No, but exploitation of the type of man that thinks the stripper is actually into him is not among the most pressing issues surrounding sex work. It's an extremely strange thing to focus on. It's like focusing on anti-white racism. Sure it exists, and it is bad. But it has nowhere near the consequences of anti-black racism, so there is no equivalence between the problems.

They are giving quite a lot permanently. Their images will always be out there on the internet and in the world, and could be used against them when applying for future jobs, etc. The sex worker generally doesn't have XXX footage of their clients. It's the other way around.

Sure, some people give them more than they can afford. Sometimes car dealerships sell people dumb cars they can't afford too.

edit: Regarding pimps, there's plenty of women who are forced by pimps to perform online. It's extremely common. I imagine moreso with COVID.

Sex workers are exploited much more than they exploit their clients. I think most people know what they are getting into.

It's illegal to use in public as others have said, but enforcing this rule isn't exactly a high priority for police. I think it's really a situation where somebody has to ask the smokers to stop, like many other obnoxious but not felonious behaviors.

Drug prohibition benefits established power structures. It's the same reason why neither major party wants to enact universal healthcare, despite being generally acknowledged as more efficient in terms of care for the cost.

I honestly don't see a huge value proposition in Scala, now that Java has reduced pain points around FP. Scala has higher-kinded types, but you see people tie themselves in knots with those as often as not. It is just such a huge language.

Scala 3/Dotty sounds like it will be a big improvement.

It doesn't have a real module system.

You won't be able to type them in Java, at least not now; however it would allow new JVM languages to do it. You could also do fibers on the JVM before, but they didn't have access to the actual JVM stack. That kind of limits their usefulness and performance.

I don't see it used much in legacy systems, but the benefits are large enough that I could see new Akka-ish libraries built on it.

Project Loom[0] is going to be coming out at some point. That brings direct JVM support for delimited continuations and fibers.

That's really going to change and simplify JVM concurrency and I think many other things. Delimited continuations can be used to implement algebraic effects, which is exciting for functional programming.

[0] https://openjdk.java.net/projects/loom/

There's no such thing as absolute proof or "knowing", only degrees of belief. Yet we make decisions based on degrees of belief all the time.

There is very good reason to believe leafy vegetables are healthy based on oodles of medical evidence spanning decades.

The 2L of water never had evidence behind specifically that quantity, and was only a reasonable general guideline for men, with it well known that the requirements vary enormously from person to person.

edit: To put it better, 2L of water a day as a good goal for the average person to shoot for is a public health message, it's not a clinical statement. There is a difference, in that public health advice needs to be easily digestible and not confuse people, even if it is leaving out nuance.

From the Mayo Clinic's site[0]:

You've probably heard the advice to drink eight glasses of water a day. That's easy to remember, and it's a reasonable goal.

Most healthy people can stay hydrated by drinking water and other fluids whenever they feel thirsty. For some people, fewer than eight glasses a day might be enough. But other people might need more.

[0] https://www.mayoclinic.org/healthy-lifestyle/nutrition-and-h...

The human body is incredibly complicated and we really don't know what we don't know. It's really not necessarily true that drinking a nutrient rich slurry like Soylent is a good substitute for a conventional healthy diet. It literally is an unknown. There's good prior reason to believe it is not significantly better. It probably isn't a whole lot worse either, certainly better than many people's diets.

Thus, unknown benefit or harm which probably isn't huge either way. It's like taking an untested drug that probably isn't dangerous but probably doesn't have a huge effect either. The precautionary principle is the usual rule with such things.

To really know, we'd need multi-decade longitudinal studies of large numbers of people. Soylent obviously doesn't have that kind of evidence.

Corporate green-washing initiatives like this just make our economic system look good, while doing no meaningful good. These corporations have political power, but they don't use it to solve this crisis.

There is a guaranteed solution to stop it from becoming apocalyptic which is comparatively easy and cheap, and we know it would work: just move away from fossil fuels and carbon intensive production ASAP. This requires political will we don't have, and fantasies like this just distract from what is necessary and the steps we must take.

These kind of pie-in-the-sky things takes money and attention from real solutions that work. They are false hopes that make people complacent and the problem seem less urgent, thinking that we'll tech our way out without fundamental changes to how we live on this planet.