wjn0

I'm not sure if this is it, but assuming the correctness of the test is independent of the sample quality, a sample could be split and tested multiple times to obtain a result with a much better FPR.

I think this might be a bit premature. For the bulk of these tools (pandas/numpy/tensorflow/torch), most of their users have probably been using them for 3-5 years max.

A universal API standard is a nice idea, but I don't know that the ecosystem has been around long enough to justify this effort.

Given the asynchronous nature of consortiums though, maybe this exactly the right time to start talking about this.

Will be watching closely either way.

Without some form of compression, I don't think it is.

In particular, consider your example (5 horcruxes with 3 needed to reconstruct). View the original file as the interval (0, N) and view it as a set covering problem. If each horcrux covers an interval of size N/3, then if any pair overlaps, there is no third horcrux that can complete the covering. This is a contradiction because 5 horcruxes of size N/3 must overlap somewhere.

> possession of 12 grams of LSD

Reader's note: that's between 50k and 100k doses.

Not that I disagree with your point.

Surely you mean the number of publications, or some other superficial metric (e.g. being the third "supervising" author on a paper with a dozen authors).

If I read a paper and have a great appreciation for the ideas, and have a sense that an author contributed significantly to the aspects I appreciate (either from explicit descriptions of the authors' contributions in the publication, or by speaking to them), why would that not be impressive to me?

> * Public access to research results, methods, and data, with some exceptions (such as PII)

You mentioned PII, so I'm assuming some familiarity with the health field. I'm curious about your thoughts on the position that one should not be required to immediately publicize their data, because there needs to be an expectation that a researcher can translate the capital (both time and money) they expend to acquire quality data into academic and institutional capital (in the form of research output, i.e. papers). The fear being, there might be insufficient motivation to conduct large data collection-oriented studies due to another researcher beating the data collector to the punch in terms of publishing certain findings.

Agree, I don't think salary caps have much to do with the poor alignment of incentives under discussion here.

That's an interesting idea that could stop very specific types of fraud, certainly in the life sciences. But it's not feasible for all kinds of research, and in fact could hinder lots of research.

> All papers would be required to be made public.

This is more universally feasible. Publicizing the data and analysis tools (scripts, software) falls into the same category, and would go a long way to help without the need for such strong separation.

Very neat. It sounds like they're curating a dataset ripe for machine learning.

Sorry, I didn't mean to detract from that point. I completely agree.

I'm not saying the policy is worthwhile or even justifiable, but the comment I replied to later used the word "treason" (sorry, I didn't quote it) which implies that these actions aren't endorsed, and certainly not encouraged, by the US government.

> Don't you think that C-levels of all of them were previously dying to get into China, and were ready to do anything?

> ...

> treason

In the case of China, doesn't the US government in some ways encourage this? I forget the colloquial name for the "policy" (if you can call it that, because I don't think it's codified into law), but isn't the idea that if the US shares some of the fruits of capitalism and liberal democracy with China, the people of their country may follow suit?

Influenza viruses are not coronaviruses.