darawk

The effect size was about a third of a standard deviation, which isn't huge, but isn't trivial either. Assuming it's true and genetically determined, that could have significant effects at the aggregate level.

However, a ray of hope in the study is that, even if you take its confidence intervals at face value (I don't), then there is quite a bit of room for improvement in the nurture sphere, as the black/white IQ test gap is generally much larger than this when not as carefully controlled as this study.

> That's why. Racism is something that we made up, and then tried to massage circumstances and data to "prove" it. This

I listed reasons to doubt the conclusions of the study. The study is generally well done and accounts for just about everything you can (practically) account for. However, it is not without flaws. It is not dispositive. But it is evidence in the direction of what's sometimes called "scientific racism".

That doesn't mean it's true. The jury is still very much out on the question. But what I was asking the GP was not "why are you unsure what's true here", but rather, "why are you so confident that one answer is true over the other".

> I also put a lie to it with my existence.

This is a misunderstanding of statistics. What this study appears to show, if you take it at face value, is that the black/white IQ gap is about 1/3 of a standard deviation. That means there are still plenty of very smart black people, and very dumb white people. It's a statement about population level averages, not individuals.

That being said, I don't take the study at face value. I think there are important confounds and the answer as to the direction and magnitude of the "genetic IQ gap" if any, is still unclear. However, what I definitely do not think is warranted by the evidence is the highly confident statement that the gap is 0.

> Sidenote: "The best study I'm aware of on the topic" is not a convincing argument that it is, in fact, even a serviceable study on the topic. However, it is a study that is often cited as such by the aforementioned bad actors. As with much racist rhetoric (spoken by avowed racists and non-racists alike), it says more about the person who brought it up than anyone else. You seem to be encountering the infohazard I mentioned earlier.

I would sincerely love for you to make me aware of more and better studies on the topic. It's something about which I'm curious, and would relish the opportunity to learn more.

That is true, and an important criticism of IQ as a measure in general. However, it's not clear that the Flynn effect is relevant to the conclusions of this study. Why do you think it is?

Why are you so confident that it's not true? The best study I'm aware of on the topic is this one:

https://en.wikipedia.org/wiki/Minnesota_Transracial_Adoption...

This concludes that there is 3-5 point genetic IQ gap between blacks and whites. This study is imperfect in a lot of ways, and fundamentally and irreconcilably confounded by the fact that even though the kids grow up in different environments raised by different families, there is no way to blind their childhood experience from the color of their own skin.

So, in no way do I mean to argue that this study is dispositive in favor of scientific racism. However, I do mean to argue that the existing state of the evidence points in that direction, albeit rather weakly.

This is a fun read and nice taxonomy that I largely agree with...right up until the end when he seemed to want to make a separate category for online warriors that say things he agrees with more frequently. It's amazing how blinding intellectual pre-commitment can be.

So much so I almost wonder if it's an intentional joke? Especially since the section was so under-explained? If so, it's brilliant.

Are you saying the demonstrated results are all in sample? Because this is definitely not true for out of sample data. And the GP comment implies that there is in fact a validation/holdout set.

Who is holding loan officers accountable?

You want to make loans without predictive optimization? Good luck with that. This is truly one of the most childish criticisms I've ever seen published.

Anyone who works on forecasting or "predictive optimization" understands all of these issues. The question isn't "is predictive optimization perfect", the question is "Is predictive optimization better or worse than your practical alternatives?" And sometimes the answer to that question is yes, and sometimes the answer is no.

It's worth reading the response to Haidt and his coauthors re-analysis:

https://sci-hub.se/https://www.nature.com/articles/s41562-02...

I haven't read all of Haidt's position yet, or the original paper, but this response is fairly convincing.

Most children didn't eat nearly as well as they do today before capitalism.

There is nothing illegal about looking like a scam. It is only illegal to be a scam.

I think the market making profits from Tether have been pretty substantial over the years. They charge very high redemption fees, which means that either they, or some close partner of theirs is the primary liquidity provider. Tether is usually mostly on peg, but it is frequently off by a few basis points, sometimes a few tens of basis points, and occasionally a percent or two. The liquidity at these levels is often quite deep, meaning you can earn fairly substantial profits by buying it up.

It's not just the headline moments when it's off peg, it's the daily liquidity churn of people willing to sell a few basis points below par, because they'd rather get out. Look at this price chart:

https://coinmarketcap.com/currencies/tether/

And look at the daily volume Tether experiences. During 2021 it traded roughly ~$100B in volume per day. One basis point on that is $10M/day. Of course, they're not a counterparty to every Tether trade, not even close, but even a fraction of that is very good money, and remember it was frequently off peg by more than one basis point. Also keep in mind the point of comparison here, which is essentially short term treasury rates over the last several years. I think the MM profit on Tether has certainly been at least close enough to match if not significantly exceed the yield on cash equivalents over Tether's lifetime (until very recently).

Note that Tether is also frequently off peg in the other direction, and they can exploit that too by minting and selling USDT into the market when that occurs. Of course, that has nothing to do with them looking dodgy or not (it's correlated to sharp, unexpected liquidity demands by retail).

> I don't think they're playing with the trust of their own stablecoin to get cheap USDT, it'd be better for them to be more trustworthy and have more USD to invest in safe investments with low interests, but they're probably losing some market share to more trustworthy (US-based) stablecoins because of how shady they look.

I think as of today that is a true statement, because yields on safe investments are now so high. But for the last decade I think it has been quite false. If Tether had only been investing in treasuries and had no other revenue strategy for the past 10 years, their profits would not be nearly as high as if they had pursued the strategy I suggested. And remember, for many years they were literally the only game in town, they had no competition to worry about.

Even in the event that Coinbase became insolvent, theoretically that shouldn't impact USDC. USDC has entirely separate reserves that are in short term debt instruments (i.e. T-bills) and cash equivalents, uncorrelated to the crypto ecosystem, and very safe.

Even if the parent company of USDC (which is actually Circle, not Coinbase) were to become insolvent, that theoretically doesn't impact USDC at all, other than maybe USDC would have to wind down operations, and then people would just redeem for USD.

The only scenario where USDC holders would be in trouble is if Circle violated their agreements and started making unsafe self-dealing investments with the USDC deposited funds. That is something that theoretically could happen, but since the operators are in the US, they would certainly go to prison for it. And we have reasonably reliable audits that state clearly that, at least so far, this has not happened.

> This doesn't seem like a particularly great strategy, compared with the alternative of running a stablecoin and exchange that doesn't look dodgy (or indeed an actual fraud which is at least insanely lucrative).

It is very much not insanely lucrative to run a normal stablecoin business. At least, not until very recently. Cash equivalent yields have been extremely low for the past decade, until the last year. USDC's net income numbers are public, and they suck.

All the people mindlessly saying "counterparty risk" clearly didn't read the article. He has thought about this carefully. His counterparty in this case is Aave, a DeFi smart contract. Smart contracts famously get hacked frequently, of course, but Aave has been around a long time, and it's probably reasonable to be somewhat confident in its security, at this point.

The second thing you have to worry about is Aave's liquidation mechanism: under the market conditions implied by a Tether collapse, will Aave's liquidation mechanisms function efficiently and effectively? The answer to that question would depend on exactly how the collapse unfolded (i.e. how quickly it was certain, the degree of insolvency, how much the market moved how fast, etc). However, it is important to note that the only thing at risk here for him is the profit from his short plus the collateral factor haircut, not the entire principal. The trade he did was to deposit $x usdc on Aave, borrow $y usdt, and then sell that usdt back to usdc. That means he physically has CF * principal USDC in his possession, and no matter what happens, Aave can't take that away from him. Now, he looped this twice, so it's actually CF^2 * USDC that he has, but that's still not that big a risk.

Finally, he has to worry about the solvency of USDC. However, USDC is regulated in the US and has fairly real audits. Almost nobody seriously thinks USDC is insolvent. I think there is very little to worry about here.

Personally, I think Tether is pretty obviously at least mostly solvent, and I think shorting it is a dumb trade that will lose him money. But he's going to lose money paying the interest, not losing his principal. People have been predicting a Tether collapse for literally years now, and despite all the market stress and volatility which should have clearly exposed their supposed fraud, they're still standing, and the peg trades with solid liquidity at $1 today.

My own theory of what's actually going on here is that Tether is intentionally obtuse, because it allows them to make seignorage profits against their own users. If you issue a stablecoin and you know you are solvent, then you can hint to the market that maybe you're not, and buy (your own) assets that you know are worth $1 at a discount, making a tidy profit in the process. I think this is their real strategy, always has been, and they've gotten very rich doing it. It's possible they've been under-collateralized at various times, and maybe are even slightly so now, but I seriously doubt they are currently insolvent to the degree people like this think.

> Why would I ever let bing crawl my site if they aren't going to send any visitors to me?

I don't think it's up to you, legally speaking: https://en.wikipedia.org/wiki/HiQ_Labs_v._LinkedIn

I mean, they could be nice and respect your robots.txt, but they certainly don't have to.

> fair use of snippets has relied on them being brief and linking to the source. Lawsuits will be immediate.

It's possible that fair use law will be expanded to cover this case, but as constructed the output of these models is generally fairly derivative of any specific original, and so probably protected under fair use. If it were spitting out exact copies of things it had read, it would probably be pretty easy to train that behavior out of it.

> I do love these imaginary scenarios where ChatGPT is going to find me the best air fryer, though. Where is that information going to come from, exactly? Barely anyone is making money writing reviews today, it's mostly farmed content. What happens when even those sites' reviews are quickly scraped and put into the next model iteration? Bing is going to have to come up with some kind of radical revenue sharing too if they want anything fresh.

I do agree with this, though. The LLMification of search is going to squeeze revenue for content creators of all kinds to literally nothing, at least if that content isn't paywalled. Which probably means that that's exactly where we're headed.

Tensors. Years ago, I would read the Wikipedia page about them every so often, and completely fail to understand what they were.

Then one day I was modeling something in a spreadsheet, and I thought to myself "you know, what I really need here, is 3rd axis to this spreadsheet". And for a few minutes, I thought I had invented a new form of spreadsheet/structure, and was considering trying to build a 3d spreadsheet program.

But, of course, suddenly everything I'd read about tensors and not understood before snapped into place, and I realized that this is exactly what they are (to be precise, this is a special case of a tensor, the contemplation of which caused me to understand what the general case is, and why e.g. a scalar is a rank 0 tensor, a vector is rank 1, a matrix is rank 2 my 3d spreadsheet idea is rank 3, etc).

But not in effort to benefit ratio, excluding the downsides of nicotine. Chewing nicotine is cheap and easy, cognitive behavioral therapy, exercise, and meditation, all take significant amounts of time at the very least.

Not arguing for the use of nicotine, but there are reasons people like taking pills (or chewing gum, as the case may be).

Where was the original thread on this? I'm not sure how I missed it. Would have been (and still am) happy to contribute.

That's exactly my point though. Figuring out the truth is very hard, and often our best experts are wrong about extremely important things. If our best experts can't get such critical things right, who can we possibly put in charge of censoring misinformation?

I don't necessarily mean to indict the scientists of Semmelweis's day too much. I mean, they fucked up, to be sure. However it's understandable (somewhat) why they didn't listen to him.

But a society that allows those presently in power to censor information they consider to be false is a society that helps cement its present ways of doing things for eternity. I don't think that's a good thing.

A modern example of this might be the plaque theory of Alzheimer's disease. I don't know enough to know whether this is a fair description of the state of the science, but there is certainly a growing view that, a "cabal" of believers in this theory suppressed alternative views for a long time. It may end up being that the cabal is right! But it seems to me like there's enough uncertainty that a diversity of viewpoints should probably be heard.

Semmelweis's lies demonstrably (according to contemporary experts) hurt the reputation of the entire medical establishment of his day. Fortunately, back then, they had the recourse to put him in an insane asylum.

Scientists and doctors regularly disagree with each other about which treatments are best for patient, how much heavily to regulate pharmaceuticals, etc. And who is right in those debates have far reaching consequences for human life. Thousands of people live or die every year based on those disagreements.

Ignaz Semmelweiss's life is the perfect example of this phenomenon in action, with extreme consequences:

https://en.wikipedia.org/wiki/Ignaz_Semmelweis

He realized doctors were killing their patients by not washing their hands and cleaning their instruments. He devised simple solutions to this problem. The doctors of his day laughed him out of the room and had him committed to an insane asylum where he was effectively tortured to death.

The consensus of medical practitioners at the time was that he was a crank. But he was right. Should he have been imprisoned? Certainly the doctors of his day would have had him banned from social media for "misinformation" if they could have, and they did far worse.

We abide by the principle of free expression not because it causes the least harm in every instance, but because it causes the least harm in expectation. The value of minority viewpoints that are right about important things that the majority is wrong about will always outweigh the harm of harmful speech, in aggregate. And one thing is clear from all of human history: We are not responsible enough to tell the difference ex ante.

1. You sort of joined two separate things here, so i'll split them:

  > Repercussions for harmful or bigoted speech, as well as escalation of responses. For example what if a person or group is being bullied but it’s claimed under free expression?

People are mostly capable of distinguishing bullying from legitimate expression. There are a few edge cases (e.g. misgendering), but they are not imo sufficiently important to throw out the basic principle of free expression.

> What happens when things escalate between parties beyond just words but physical harm?

I don't really understand why this is even a question? Physical assault is not only contrary to MIT's rules, it is a crime. This statement and principle is totally orthogonal to the principle of free expression.

EDIT: I suppose you could mean something more like incitement/inducement to physical harm. But, of course, that is also a crime.

> 2. Similarly how do they stop majority dominance and snowballing into group think? E.g if there are way more men or women in a program and they use their larger numbers to amplify their freedom of expression in a way that affects the others. That then risks reducing the number of the minority demographic as they slowly get pushed out, leading to largely heterogeneous programs.

This is a strange view to me. Is the underlying premise here that there are things that the majority believes, but that would be harmful to say? If so, how did we come to this conclusion and what strange things could be believed by the majority of students/faculty but are too dangerous to speak? Certainly we could point to historical examples of things people in say, the 1950s, believed that we might consider dangerous or harmful to say today. But people from the 1950s would never have chosen to ban those ideas.

In other words, nearly by definition, it is never the preferred expressions of the majority that are getting regulated. If it were a majority view, it wouldn't be considered harmful by the regulators. The viewpoints having their speech regulated are always going to be in the minority and unpopular. If it's not a majority that believes these things, then there should be no reason to be concerned about majority dominance.

The power to regulate expression is just that - power. If you are truly concerned about concentration of power and the protection of minority interests, there is no greater threat to that than the regulation of expression. Tautologically, it will always be those in power that write the speech codes.

20% is not for possession. 20% is all drug-related charges. Many if not most of those are going to be intent to sell.

> I guess it's a question of personal interpretation - but to me those two numbers aren't different enough to quibble over.

Which two numbers? The only number I quoted is 20%, and that's for all drug inmates, not just possession. We don't have the numbers for possession, but they're surely much, much lower.

I guess my prior based on the way this guy was talking was that he probably had at least 6-7 figures in crypto stolen, which is worth it for a mid tier 0day. Especially if you think it's unlikely to actually get burnt.

Why not? I have a security background. I see nothing wrong with that statement. Although what he actually said was:

"Initially I imagined I was targeted by a 0day or rootkit"

which actually does not make sense, because it implies he thinks those two things are fungible. He's obviously not a security expert, but he's also obviously not totally technically incompetent.

This is a common myth. The percentage of people locked up in the US for drug offenses in general is only 20%. The share of those there specifically for possession is much lower still:

https://www.prisonpolicy.org/reports/pie2022.html

Police do make a lot of drug possession arrests, but the actual prison population who are there for possession is nowhere near 33%.

It seems like a reasonably well written anecdote by someone who has some idea what they're talking about. It could obviously be false, but the consequences if he's right are potentially serious for a lot of HN users who might use LastPass. The consequences if he's wrong are a little extra reputational damage for LastPass, but that seems like a worthwhile tradeoff here.

Not everything posted on HN has to be verified true. The decision calculus here seems strongly in favor of signal boosting it, so that people who need to can take defensive action, even if it turns out to be wrong.

This is an excellent article. Something bugged me about the "adult content spam" explanation for the Twitter issue when it circulated, but I couldn't put my figure on what it was. It just didn't feel like that story really held together for some reason. But this explanation totally makes sense, and is an important fact to keep in mind when evaluating any claims of this sort.