Douglas_Knight

Karma: 7,637

Douglas_Knight 14 May 2014 18:50 UTC
55 points
on: [LINK] Sentient Robots Not Possible According To Math Proof

These authors argue that consciousness should be based on lossless integration, otherwise the memory traces would be affected each time we remember something. Problem is, this is exactly what happens.

- Jay

Douglas_Knight 20 Nov 2012 23:37 UTC
50 points
on: Miracle Mineral Supplement
I don’t think point and sputter posts like this are very useful. How is this example more surprising than any other quack medicine example? How much understanding does the typical patient have of any medicine? Lots of medicines are controlled doses of poisons.

I think you are mistaken about the lethality. It would be surprising if it “often” killed its users, yet was able to spread. But that’s not true. Yes, it is sold at industrial concentrations, but most people follow the directions and dilute it. The FSA says that, used as directed, it will only cause GI distress (though the FDA suggests that the low blood pressure could be fatal). Users are warned of the effects ahead of time. That probably reassures them that it is working, that they haven’t been scammed with an inactive substance.

What is the death rate? The Seattle case appears to involve 200 users and no fatalities. The woman who died in Vanuatu appears to be the only known death, but I don’t think much is known about the hundred thousand malaria victims in East Africa who took it.

I don’t think this demonstrates human stupidity any more than any other quack medicine example. It does nicely illustrate Poe’s law. That may make it more memorable and convincing.

Added: By “point and sputter” I mean that Eliezer did not provide enough information for me to determine what surprised him about this example and why it would be a useful example to give to others. I think he reached much of his conclusions from false beliefs about how the product is described and how lethal it is, but I don’t know. Certainly, what the impression I took away from his post was false.

Douglas_Knight 26 Mar 2024 19:42 UTC
43 points
36
on: My Interview With Cade Metz on His Reporting About Slate Star Codex
That they have a “real names” policy is a blatant lie.

They withhold “real names” every day, even ones so “obvious” as to be in wikipedia. If they hate the subject, such as Virgil Texas, they assert that it is a pen name. Their treatment of Scott is off the charts hostile.

If we cannot achieve common knowledge of this, what is the point of any other detail?

Douglas_Knight 17 Jul 2011 17:23 UTC
41 points
on: How to solve the national debt deadlock

The US Congress is trying to resolve the national debt

That’s just not true.
What links here?
- RobertLumley's comment on Please do not downvote every comment or post someone has ever made as a retaliation tactic. by Will_Newsome (21 Aug 2011 13:26 UTC; 10 points)

Douglas_Knight 29 Mar 2023 18:47 UTC
40 points
5
on: Which parts of the existing internet are already likely to be in (GPT-5/other soon-to-be-trained LLMs)’s training corpus?
I assume you know this, but to be clear, OpenAI has already used pirated books. GPT-3 was trained on “books2” which appears to be all the text on libgen (and pretty much all the books on libgen have been through OCR). It was weighted the same as the common crawl, lower than Gutenberg or Reddit links. This seems to answer your second question: they will likely treat pdfs on the libgen the same as pdfs on the open web. If you’re asking about whether they will train the model on the pixels in these pdfs, which might make up for losses in OCR, I have no idea.

Douglas_Knight 1 Feb 2010 2:56 UTC
38 points
on: Deontology for Consequentialists
Sometimes I believe that
- consequentialism calls possible worlds good
- deontology calls acts good
- virtue ethics calls people good
Of course, everyone uses “good” to label all three, but the difference is what is fundamental. cf Richard Chappell

Douglas_Knight 15 Feb 2015 23:02 UTC
35 points
on: The Galileo affair: who was on the side of rationality?
Yes, Galileo was an asshole. Yes, he had many bad arguments and couldn’t tell which were better than others. Yes, Riccioli was a good scientist. But the conventional wisdom is pretty much correct. Your essay is largely cherry-picked and misleading. You may not be Catholic, but I’ve heard identical ones repeatedly from Catholics, including identical ignorance of what Galileo actually contributed. So it appears to me that your sources were designed for the purpose of rehabilitating the Inquisition.

Also, the question is not merely, whether the weight of the evidence in 1616 or 1633 was against of heliocentrism, but whether it was so overwhelmingly in favor that argument should be banned.

his opponents were citing Aristotle, not the Bible in their arguments

They were citing both. Most of them did not consider the Bible infallible on temporal matters, but they did consider it powerful evidence, starting with Martin Luther, who condemned Copernicus by citing Joshua’s miracle of stopping the motion of the sun across the sky.

Martin Luther may seem out of place in Catholic arguments, but he changed history. His contemporary Copernicus didn’t run into trouble because the Catholic Church was secure. He died before the Counter-Reformation.

Last time I looked into this, I concluded that the Inquisition (which is not just Riccioli!) couldn’t even tell which hypotheses were being argued about. There are really three hypotheses: stationary geocentrism, rotating geocentrism, and heliocentrism. The hypothesis that the sun and planets revolved about the Earth, but that the Earth rotated on its axis every day had many followers, both in antiquity and the Renaissance, yet few noticed that already this theory, and not just full-blown heliocentrism contradicts Joshua. Galileo’s greatest contribution was Galilean relativity, arguments about why we would not notice the rotation of hundreds of meters per second.

The heliocentric view had only a single advantage against the geocentric one: it could describe the motion of the planets by a much simper formula.

That’s a pretty big advantage. But Galileo provided a lot more. He famously observed the moons of Jupiter. The really big contribution of the telescope was the phases of Venus, which really trumps anything else. (You could take the position that only Venus and Mercury went around the sun. This was occasionally put forward in both Antiquity and the Renaissance. That it never took hold, even before Galileo, I find bewildering. It makes it difficult to judge anything else.)
What links here?
- komponisto's comment on The Galileo affair: who was on the side of rationality? by Val (17 Feb 2015 10:28 UTC; 0 points)

The Cold War divided Science

Douglas_Knight5 Apr 2014 23:10 UTC

34 points

53 comments1 min readLW link

Douglas_Knight 30 Aug 2012 20:51 UTC
33 points
in reply to: DanArmak’s comment on: [META] Karma for last 30 days?
Also, visible actions spread. Someone who posts a correction encourages other people to post corrections to other comments, while someone who PMs does not encourage that behavior. But it is also visible how many responses there are, so that people don’t overwhelm an individual with responses, while they might overwhelm with PMs.

Douglas_Knight 1 May 2013 22:11 UTC
32 points
on: What do professional philosophers believe, and why?
Principal component analysis is not the same as clustering. Some of the post seems to make that distinction, while other parts appear to blur it.

If there are clusters, PCA might find them, but PCA might tell you something interesting even if there are no clusters. But if there are clusters, the factors that PCA finds won’t be the clusters, but the differences between them. In the simplest case, if there are two clusters, a left-wing party and a right-wing party, PCA will say that there is one interesting factor, the factor that distinguishes the two parties. But PCA will also say that is an interesting factor if there is no cluster, but a clump of people in the middle, thinning towards extremists.

Actually, factor analysis pretty much assumes that there aren’t clusters. If factor 1 put you in a cluster, that would tell pretty much all there is to say and would pin down your factor 2, but the idea in factor analysis is that your factor 2 is designed to be as free as possible, despite knowing factor 1.
What links here?
- What do professional philosophers believe, and why? by Rob Bensinger (1 May 2013 14:40 UTC; 49 points)
- badger's comment on What do professional philosophers believe, and why? by Rob Bensinger (1 May 2013 23:30 UTC; 8 points)

Douglas_Knight 4 Mar 2012 3:46 UTC
32 points
in reply to: Scott Alexander’s comment on: How to Fix Science
Yes, “four times as likely” is not the same as an odds ratio of four. And the problem here is the same as the problem in army1987′s LL link that odds ratios get mangled in transmission.

But I like odds ratios. In the limit of small probability, odds ratios are the same as “times as likely.” But there’s nothing 4x as likely as 50%. Does that mean that 50% is very similar to all larger probabilities? Odds ratios are unchanged (or inverted) by taking complements: 4% to 1% is an odds ratio of about 4; 99% to 96% is also 4 (actually 4.1 in both cases). Complementation is exactly what’s going on here. The drug companies get 1.2x-1.3x more positive results than the independent studies. That doesn’t sound so big, but everyone is likely to get positive results. If we speak in terms of negative results, the independent studies are 2-3x likely to get negative results as the drug companies. Now it sounds like a big effect.

Odds ratios give a canonical distance between probabilities that doesn’t let people cherry-pick between 34% more positives and 3x more negatives. They give us a way to compare any two probabilities that is the obvious one for very small probabilities and is related to the obvious one for very large probabilities. The cost of interpolating between the ends is that they are confusing in the middle. In particular, this “3x more negatives” turns into an odds ratio of 4.

Sometimes 50% really is similar to all larger probabilities. Sometimes you have a specific view on things and should use that, rather than the off the shelf odd ratio. But that doesn’t seem to be true here.

Douglas_Knight 6 Jan 2015 3:25 UTC
29 points
on: Negative polyamory outcomes?
In my experience people who have recently gotten into poly are evangelical, claiming that it is easy and for everyone, while people have done it for years say it is hard and not for everyone. But you hear a lot more from the people with the new partner infatuation. Experienced people complain even about this dynamic.

Douglas_Knight 24 Nov 2013 6:57 UTC
27 points
on: On Walmart, And Who Bears Responsibility For the Poor
Your analysis of the short-term effects is correct, but the long term effects depend on whether “low wage workers” are permanently so. Sometimes people condemn Walmart jobs as “dead-end” and that is getting at the right point.

I’ve heard the claim that Costo and Sam’s Club (ie, Walmart) are very similar, but Costco is famous for paying its employees twice as much. But this doesn’t come out of profits—Costco spends the same amount on labor, employing half as many people, twice as productive. If Walmart could make its employees twice as productive, that would be great for society, though in the short term it would lay off half of them.

If the productivity of people is unchangeable, then Walmart is doing society a valuable service by providing a niche to people capable of no more. But if Costco employees are more productive because Costco trains them, then Costco is doing a valuable service by improving their productivity. In the first case, we want Walmart to win because only a few companies like Walmart can make use of the least productive workers. But in the second case, we want Costco to win because it is making use of the same people, but making better use. But we observe that they are evenly matched, so there’s no reason to expect either of them to win, let alone the right one. Eventually in the second scenario Walmart loses, not because Costco wins, but only when the Costco model expands into new industries, producing more training, bidding up the salaries Walmart pays.

In the particular example, I believe that Costco is not increasing productivity, but merely identifying more productive workers, and that Walmart is able to employ people that few other companies can. In general, I think the economy is generally trending away from investing in low-end worker productivity, which is terrible. In theory, raising the minimum wage should put pressure against this, but pressure to create new companies that work differently is less certain than pressure shifting the balance of power between existing companies.

(Also, there’s a third scenario where Walmart provides the training, but the productive workers graduate to Costco. I certainly think Walmart is providing filtering, letting productive people build up a resume to show to Costco; I’m less certain of whether it improves the workers.)

Douglas_Knight 24 Jun 2012 15:14 UTC
26 points
on: Minimum Viable Workout Routine is Dangerously Misinformative
As a general rule, it seems to me that if you’re going to call something dangerous, it behooves you to spell out exactly what it is you think it is claiming and exactly how you disagree with it. I see very little disagreement between the two posts and no even attempt to justify the word “dangerous.”

Douglas_Knight 3 Feb 2012 5:59 UTC
26 points
on: Fireplace Delusions [LINK]
I call bullshit.

He doesn’t quantify anything sensibly. Sure, maybe wood smoke is 30x more carcinogenic than cigarette smoke, but they are consumed very differently. Smokers stood out by getting the rare disease of lung cancer at a time when wood heating was quite common. Maybe it’s comparable to living with a smoker, but he doesn’t say. Sure, it’s bad for you, but so is skiing. Is it worth it as recreation? Nothing he says addresses this. I would guess that the life expectancy cost from the dangers he mentions is less than the expected cost from the acute danger of messing up the flue, passing out from CO and dieing (something that nearly happened to my mother).

(When he talks about burning solid fuel in the developing world, does that include coal?)

Douglas_Knight 7 Sep 2022 16:57 UTC
23 points
9
on: Let’s Terraform West Texas
I would distinguish terraforming from irrigation. It sounds like you are talking about setting up a self-contained system to irrigate the land every year, whereas I would restrict terraforming to a permanent climate change, so that it rains and desalination is no longer necessary. This is what people mean when they talk about terraforming the Sahara. The desert was green five thousand years ago when the pyramids were built, so it probably has multiple equilibrium climates and a sufficient intervention could get it to jump to the other. I don’t know how plausible this is for other deserts. The idea is to irrigate a bounded number of times to grow appropriate trees to trap water.
A permanent change could be much cheaper per acre because the solar panels and desalination plant can be reused for new parcels. The downside is that this probably has large gains from scale: air flows freely between neighboring parcels and thus humidity is an externality. Whereas the whole point of a self-contained system is that you can start small.

Douglas_Knight 15 May 2013 3:02 UTC
23 points
in reply to: DanielLC’s comment on: Avoiding the emergency room
This is pretty standard advice from police and is probably not based on emergency room experience. The advice is to obey people who threaten immediate harm, except to refuse to go anywhere. The claim is that the main reason for transit is to go somewhere safer for the assailant, compared to shooting you on the street.

Douglas_Knight 23 Feb 2012 1:35 UTC
23 points
in reply to: Vladimir_Nesov’s comment on: [link] Faster than light neutrinos due to loose fiber optic cable.
By making this complaint now and not earlier, you are promoting publication bias.

Douglas_Knight 15 Feb 2023 18:19 UTC
22 points
5
on: Bing Chat is blatantly, aggressively misaligned
In what sense are these failures agentic?

Douglas_Knight 11 Jan 2015 19:13 UTC
22 points
on: How Islamic terrorists reduced terrorism in the US
I know you didn’t make the graphs, but …

I don’t like line graphs. They’re a kind of smoothing, but almost never the right choice. The first two graphs are smooth enough that it hardly matters what you do, but I redid the third as a scatterplot plus a loess curve.

Here’s the loess superimposed on the line graph, if you want to compare them. I got the data here, linked from here

Douglas_Knight

The Cold War di­vided Science

The Cold War divided Science