fastaguy88

The idea that "courts" are equipped to address scientific questions is ridiculous. Perhaps French courts are better -- in the US, it is very much about decisions based on a confrontational presentation of "evidence", something like "debates" about evolution or vaccine injuries.

Debates are about persuasion, not reality. And, in some contexts (like whether you should go to jail), there is a process argument that there needs to be a debate-like process. Science deals in facts and statistics that have associated uncertainty. For scientific questions, there should be better decision making processes than courts.

There is a fixed (decreasing) budget for research.

What should be the balance between repeating things that have already been done, and expanding knowledge?

I suppose "totally unrelated" might mean not in the same genus to some set of taxonomists, but for the larger audience to which this essay was directed, unrelated might mean "is not related in any way" (as in, does not share a common ancestor).

Yes, they are "totally unrelated" in the sense that I am likely totally unrelated to you, but this not a very evolutionary sense of the word "related".

Two nits to pick, one major, one minor:

Major:

   "Common names simply don’t work on a large scale. One common name (ie    “cedar”) can refer to 8 different totally unrelated plants, where as Cedrus refers specifically to the genus which contains the species C. libani, C. deodara,  and C. atlantica."

Actually, those 8 different plants are hardly "totally unrelated". They are quite closely related in the evolutionary scheme of things (plants have been around for 1.6 billion years). Red cedar (Juniperus virginiana) and Cedrus (Cedrus atlantica) are both plants, but also Pinidae (conifers), and diverged about 250 Million years ago (time tree.org), so more distantly related that mammals (~80 y), but midway between humans/platypus (200 My) and humans/ducks (320 My). Hardly unrelated.

Minor:

   "Why is it [taxonomy] cool? Because we now group things according to how evolutionarily related are, and how they evolved."

Yes, we group things by how closely related they are, but not necessarily by how distantly related they are in time. Taxonomy is about branch order, not distance in evolutionary time. The same species in a genus may include organisms that diverged several million years ago, but other species may have diverged much more recently. Taxonomy is a kind of evolutionary history that need not have a very regular clock.

Did I miss the part where you must be looking at your iPhone for it to work?

Download for mac. Find out I need a different subscription. Cannot quit program (must force quit).

Perhaps I need AI to use it.

In 1998-1999, I was the organizer for a Bioinformatics course in Sao Paulo Brazil, funded in some indirect way by the National Academy of Sciences (as I recall). In addition to paying for the faculty (about 6 people) to spend a week teaching, the budget also paid to purchase an SGI server, which was a standard machine to run the bioinformatics software. The course ran for a week in January, 1999, and we had ordered (and paid for) the machine by July, 1998.

A group of us arrived in Sao Paulo on Thursday (before the course was scheduled to start on Monday) to begin configuring the machine -- downloading/transferring software and databases. We did not trust the internet, so we brought the stuff we needed on a laptop.

We got there thursday, to find out that despite a 6 month lead-time, the machine had not been delivered. Not only that, the weekend we planned to configure the machine, the power to the campus was scheduled to be shut off so that some of the 60+ year old infrastructure could be replaced (our host got a special dispensation to keep the power on in the building we were using). After many frantic calls, the machine was delivered around 5:00 pm on Friday. We worked frantically through the weekend, and managed to get a few things working, a few hours before our first computer lab. Fun times.

And yet, thousands and thousands of 'C' programs parse integers every hour successfully.

Perhaps the right title should be "No way to parse pathological edge cases in 'C'"

And then see how other languages do.

While it is true that departments often fund their graduate students for the first year (or possibly 2) out of their own budgets, their budgets are largely determined by the generosity of their Deans (who got the money from indirect costs from grants) or their own indirect costs. And they will not be admitting students if they do not see a clear path for them to be externally funded after their first year or two.

This is not true for PhD programs in top-ranked institutions. It may have been true 20+ years ago, but today it is very difficult to buy your way into a graduate program.

There are no un-funded graduate (PhD) students in the sciences and engineering at MIT (or any other top-ranked graduate program). The number of graduate student admissions is directly tied to the amount of external funding. If the faculty do not have the grants, their departments cannot admit students.

Lugaru (??) emacs (epsilon) on CP/M in the early 1980’s

BioBank claims (1) only de-identified data was available and (2) none of the data was actually sold before the datasets were taken down.

Well, perhaps Orsen Scott Card does not need editors’ advice. But odds are you do.

Amusingly, 15 min ago, the animation did not work on my MacOS Safari (Sequoia), but it was visible on Chrome. Now (1600 UTC-4), it is animated on Safari.

Apple Keychain has a number of old bugs that have caused me to have to resort to this strategy several times. The most common problem is having a secure note that you can open, but then immediately disappears (closes). Copying over an older keychain database can sometimes solve the problem.

>>> If I buy a can of soup and find glass in it, I have a valid claim against the manufacturer. It's a matter of holding someone accountable for fraud or negligence, not a matter of regulation. The proper route is a court, not a bureaucratic agency that preemptively dictates production methods on the assumption that every manufacturer is a potential prisoner.

This is a common conservative trope. We don't need regulation because customers can always sue. (Famous interview with Milton Friedman.) Good luck finding a lawyer who will sue because of some glass in your soup can, or, for more serious cases, who can out last (or match the spending of) a billion dollar corporation. Yes, sometimes the underdog wins. Rich people can sue, and may not need the governments regulatory help. For most people, there is absolutely no recourse, particular for technically complex things, like prescription drugs.

The idea that the legal system can consistently make better informed technical decisions than government scientists is not well supported by the evidence.

Re self-correcting science. In the area I am most familiar with (basic life sciences), correction happens pretty quickly. But I don’t worry about public policy much.

But I’m comfortable arguing that where science intersects with policy, fraud plays a very minor role. I suspect that most policy “mistakes” (policies that were adopted and then reversed) are more about the need for a policy in the absence of data (covid and masks), or subtle tradeoffs (covid and masks), or a policy choice that seems slightly better than an alternative (mammography) but also has poorly understood harms. Policy involves politics, and science unfortunately plays less of a role than one might like (and fraudulent science an even smaller role). This is not my field, but I cannot think of policies that were reversed because of discoveries of fraud (perhaps thalidomide and other drug approvals).

It is useful to distinguish between "effective" scientific fraud, where some set of fraudulent papers are published that drive a discipline in an unproductive direction, and "administrative" scientific fraud, where individuals use pseudo-scientific measures (H-index, rankings, etc) to make allocation decisions (grants, tenure, etc). This article suggests that administrative scientific fraud has become more accessible, but it is very unclear whether this is having a major impact on science as it is practiced.

Non-scientists often seem to think that if a paper is published, it is likely to be true. Most practicing scientists are much more skeptical. When I read a that paper sounds interesting in a high impact journal, I am constantly trying to figure out whether I should believe it. If it goes against a vast amount of science (e.g. bacteria that use arsenic rather than phosphorus in their DNA), I don't believe it (and can think of lots of ways to show that it is wrong). In lower impact journals, papers make claims that are not very surprising, so if they are fraudulent in some way, I don't care.

Science has to be reproducible, but more importantly, it must be possible to build on a set of results to extend them. Some results are hard to reproduce because the methods are technically challenging. But if results cannot be extended, they have little effect. Science really is self-correcting, and correction happens faster for results that matter. Not all fraud has the same impact. Most fraud is unfortunate, and should be reduced, but has a short lived impact.

It is perhaps worth noting that the 25,000 bit flips/out of 470,000 crashes (in a week) are probably not coming from all Firefox users. It would be useful to know how many of those crashes (and bit flips) are happening on the same machine. And whether the crashes/bit flips continue on the same machine continue from week to week.

I can certainly imagine that a very small fraction of Firefox users are generating these results, so that bit flips are not a problem generally.

Nice display, but it starts off with misleading measurements of DNA. The spacing between DNA base pairs is 0.34 nm, so the 10 base pairs pictured are in fact 3.4 nm. But the DNA in a single human cell is about 2 meters, and chromosome lengths vary from 2 to 10 cm. I am skeptical of the hemoglobin vs ribosome sizes as well; hemoglobin has a molecular weight of about 60,000, while ribosomes weigh more than 5 million.

Before Postscript, tektronix 4010 was the language of graphics, at least for biologists drawing phylogenetic trees or reassociation curves. The phylip phylogenetics package, introduced around 1980, still has the option to generate trees in tek4010 format.

You're kidding, right? (p++)[0] returns the contents of (p) before the ++. Its hard to imagine a more confusing juxtaposition.

When I think of things being in free-fall, I think of them going down. Not going up more slowly.

Sounds more like he was scammed by the dealer than a Tesla problem.

It really depends. Working on genome analysis, I once encountered/interrupted (by rebooting after a software update) a student who had been running an analysis for more than a week, because they had not pre-sorted the data. With pre-sorted data, it took a few minutes.

Not everyone works on web sites using well-optimized libraries; some people need to know about N and Nlog(N) vs N^2.

Really not a libertarian, but why shouldn’t Netflix have the right to choose who they distribute content to? They negotiated conditions with the creators, why shouldn’t they be able to specify the DRM? No one is forcing you to subscribe to Netflix. Or even to buy an iPad.

This is an interesting insight. The OP's constraint that no two adjacent squares are the same color ensures non-randomness. (Which reminds us why people are so bad at producing "random" sequences.)

PG is an excellent writer, but this essay seems remarkably misleading. The unstated premise seems to be that well-educated adults already know everything they want/need to know about everything, which is silly. I'm older than PG, and pretty well educated, but I am constantly learning new things. I don't think it's because they are not important or I am obtuse. I think it is because I am (still) intellectually curious.

Sometimes I learn new things because they are new. And sometimes I learn new things (that are well known to people in other fields) because while I know a lot about some things, I know very little about others -- so little that I don't even know those things overlap with my interests.

Those of us who enjoy learning appreciate that we will never know everything we would like to, and in fact we will never know the boundaries of knowledge for topics we care a lot about. It's not that it is unimportant to us, it's just that we hadn't learned about it yet. That's why we read essays.

In the protein annotation world, which is largely driven by inferring common ancestry between a protein of unknown function and one of known function, common error thresholds range from FDR of 0.001 to 10^-6. Even a 1% error rate would be considered abysmal. This is in part because it is trivial to get 95% accuracy in prediction; the challenging problem is to get some large fraction of the non-trivial 5% correct.

"Acceptable" thresholds are problem specific. For AI to make a meaningful contribution to protein function prediction, it must do substantially better than current methods, not just better than some arbitrary threshold.