eemax

The Illusion of Security

Flock advertises a drop in crime, but the true cost is a culture of mistrust and preemptive suspicion. As the EFF warns, communities are being sold a false promise of safety - at the expense of civil rights* (EFF).

...

True safety comes from healthy, empowered communities; not automated suspicion. Community-led safety initiatives have demonstrated significant results: North Lawndale saw a 58% decrease in gun violence after READI Chicago began implementing their program there. In cities nationwide, the presence of local nonprofits has been statistically linked to reductions in homicide, violent crime, and property crime (Brennan Center, The DePaulia, American Sociological Association).

These are incredibly weak arguments. I haven't personally looked into how good Flock cameras are at actually preventing crime and catching criminals, but if this is the best counterargument their detractors can come up with, it makes me suspect they're actually pretty good.

Crime is extremely bad. Mass surveillance is bad too, especially if abused, but being glib or dismissive about the real trade-offs is counterproductive.

Also, recording in public spaces (or private spaces that you own) is an important and fundamental right just like the right to privacy; simply banning this kind of surveillance would also infringe on civil liberties in a different way. I agree that laws and norms need adjusting in light of new technology, but that discussion needs more nuance than this.

It is true that only Yudkowsky gets to say what the rational conclusion of his ideas are. Nobody else gets to speculate. Only the pope of rationalism, because he's the rational one here. See? It's right there in the name!

No, I am saying that Yudkowsky's views are straightforwardly compatible with bedrock principles of liberalism, and the author of the piece fails to acknowledge that compatibility or grapple with them himself. It's not about "rationalism" or who is "allowed" to speculate.

I called it slop because it says false things that have the hallmark of LLM style, e.g.

The Sequences build the liturgy: a small caste of correct thinkers, epistemically and morally superior, whose rationality entitles them to govern what the rest of humanity is allowed to build. It’s not a safety movement. It’s a priesthood with an origin story written in fanfiction.

The Rational Conclusion of Doomerism Is Violence

No it isn't. The most prominent "doomer" has a strong grasp and deep, wholehearted appreciation for the the principles of liberalism and the rule of law:

https://x.com/ESYudkowsky/status/2043601524815716866

Which the author of this piece of slop appears to lack.

We pin all of our npm dependencies and upgrade them via dependabot. Dependabot links to the GitHub or GitLab release for each dependency bump, and I typically skim / scan every single commit to each dependency. But there's no guarantee that what's on GH matches what is uploaded to npm (which is what happened in this case; there are no malicious commits).

Does anyone know of a good way to verify that a npm release matches what's on GH? Version controlling the entirety of node_modules/ and running untrusted updates in a sandbox would work in theory, but in practice many packages contain minified js which makes the diffs between version bumps unreadable.

Cool post! I don't have a ton of experience using Stripe, but shouldn't you at least be handling some sort of payment_failed webhook?

It looks like you call _createSubscription and set the initial value for currentPeriodEnds before you know the payment actually succeeded, and since you don't ever check or listen for failed payments, anyone could get a free month (or year) of Checkly by using a bad card, or if the payment just randomly fails.

Maybe this isn't a huge deal in the early days, but you and your customer might not even notice the failed payment for quite a while unless you happen to check your Stripe dashboard!

The library described in this blog post looks kind of interesting as an implementation of a Maybe monad in python, but the example case is pretty silly.

It re-implements a 4 line function as a 13 line class, but the logic at the caller doesn't get any simpler:

  try:
     result = get_user_profile(id)
  except:
     # handle any exceptions...

  result = FetchUserProfile(id)
  if (result is a Failure):
    # handle the failure
  # do something with the result

A gold standard is terrible monetary policy, even worse than the status quo.

Good monetary policy would be the fed doing NGDP level targeting a la Scott Sumner: https://www.mercatus.org/system/files/NGDP_Sumner_v-10%20cop...

Monetary policy only goes so far though, if you actually want to distribute resources in an equitable way, you probably also need good tax policy.

Good tax policy would be taxes which decrease wealth inequality (e.g. luxury consumption taxes, land value taxes).

Well, to some degree, people probably do want to live in big, new houses. There's also probably some "real" underlying demand for an education, outside of people needing it (or believing that they need it) to get a good job.

But the point is, regardless of how much people "actually" want them, it is in fact technologically and ecologically possible to provide most people in the U.S. with these goods. (The evidence is that a lot of people do in fact have these things, even if on paper they are underwater on a mortgage or saddled with a big student loan.)

The bad policy is that by subsidizing and insuring these loans / mortgages far too cheaply, the government is driving up demand, and as a result people buy a bigger house than they need, or get an education that isn't useful for getting a high-paying job.

If the government stops subsidizing mortgages and student loans with bailouts, insurance, laws preventing default, etc. the result will probably be

1) the cost of a college education goes down and also that less people actually go to college 2) people don't buy as many homes and buy smaller houses than they otherwise would have

If you then want more people to have houses, education, healthcare, food, etc., whether or not they are able / willing to sell enough of their labor to pay for these things themselves, the next step is to solve wealth inequality.

We could probably fix wealth inequality with good monetary policy (e.g. NGDP targeting) and "taxing the rich" in a sane way, e.g. with land value taxes, luxury consumption, VAT taxes, etc. After that, you're probably done, but for the remaining poor, the government can just give them cash, e.g. basic income, instead of trying to subsidize a million different services like healthcare and education and homes, in which case the subsidies end up in the pockets of doctors and administrators and landlords.

Well, it does actually solve the problem in the sense that it actually causes houses to be built and students to go to college. Probably this indicates the real problem is more about bad monetary policy and bad tax policy, not a fundamental shortage of resources.

Of course, it would be better to improve monetary policy rather than just making a bunch of dumb, risky loans leading to a dramatic boom and bust.

But that would require policymakers understanding basic macroeconomics, so don't hold your breath...

Yes, this is probably effective propaganda, and moreover, propaganda for a just cause.

Nonetheless it is not evidence for climate change, and moreover it is not even necessarily an example of the effects of climate change - even we 100% fix climate change, there will still be plenty of starving wild animals available for sad videos.

The actual way to ensure there are no starving wild animals is to exterminate them in the wild. But that's probably not a good idea.

Individual animals in the wild die cruel, slow, painful deaths all the time, regardless of whether their species as a whole is thriving or not.

So while this video might evoke a strong emotional response, it is just an anecdote and not actually evidence one way or the other for human-caused climate change.

Of course, human-caused climate change is obviously real, but it's not OK to use bad arguments even if your conclusion is correct.

EDIT: To clarify further, I am not disputing that polar bears are not being harmed by climate change, or even that the particular bear in the video is not starving because of climate change. I am merely pointing out that one could produce a video of a starving polar bear both in our world and in the world where climate change is not happening, because wild animals starve all the time.

Isn't it much more likely that Tether is keeping some or most of its reserves in bitcoin, rather than USD, as they claim?

Most people get Tethers from trading on bitcoin exchanges, not from buying it with USD. If Tether accepts payment for tethers in bitcoin, they're supposed to immediately cash the bitcoins out to fiat. But my guess is they're not actually doing this, and just keeping a good portion of their reserves in bitcoin. If that's the case, they'll be solvent as long as the price of bitcoin doesn't collapse (big if) - in fact, they would have made a fortune over the last couple of months.

No, not really. It makes it useful for criminals, and a small number of people who might have good reasons to buy things, but that doesn't mean it should have a market cap of billions.

I'm not saying there are no good uses for bitcoin, of course. For example, if someone used bitcoin to import prescription drugs from Canada to the U.S, that's great. But it doesn't justify a $100B+ market cap, and as it stands, the hard part of doing that is importing the drugs, not exporting the money.

Similarly, just because Tor is useful mostly to criminals and a few human-rights advocates or privacy enthusiasts, I don't expect the Tor Browser to ever gain a major market share.

Because it seems like an obvious bubble, with little / no real value?

Bitcoin was / is a pretty cool technological innovation. But the main use-case seems to be buying illegal things or laundering money, and vast majority of transactions are just speculators.

Also, the fact that it is difficult to short is further evidence of a bubble.

I should clarify that I am in the U.S... seems like most of these sites are unavailable here.

Related to this, does anyone know of a way to bet against bitcoin (or cryptocurrencies in general), without buying cryptocurrencies themselves?

There are lots of more-or-less sketchy cryptocurrency exchanges which allow you to shortsell bitcoin, but these all require buying cryptocurrency in the first place as collateral.

The upshot is that if you bet against bitcoin and win, you win a bunch of... bitcoin. Not so useful if you want to bet on a big collapse of crypto in general.

Selling bitcoin futures seems like a way to bet against bitcoin in the long term, using actual USD, although I don't think that will ever be an option for smalltime investors / individuals.

Nice. In general, I think cryptocurrency and blockchains are pretty over-hyped, but they are a good example of a non-trivial but still simple application to show off a programming language.

For example, compare this to https://news.ycombinator.com/item?id=14439789

for a simple blockchain implementation in Haskell.

The comparisons in this article are mostly against the high-end Intel core line, but these CPUs support server / enterprise type features like ECC memory, lots of PCI-E lanes, and virtualization features (I think?).

Shouldn't Threadripper be compared to Xeons?

EDIT: Or rather, what I'm really wondering is what these CPUs lack that AMD's server line (EPYC) have.

I've never heard of Capture the Flag. Can someone a bit more familiar describe what the format is exactly, or what type of questions / challenges there are? Any examples from last year's competition?

Neat. Just browsing through the source, is there a reason that:

   calculateBlockHash (Block i p t b _)  = concatMap hashString [show i, p, show t, b]

   calculateBlockHash (Block i p t b _)  =  hashString $ concat [show i, p, show t, b]

edit: thinking about it a little more, option 2 seems better because in option 1 you can calculate 3/4 of a block hash without knowing the previous block's hash. Maybe that's not a problem in this context though?

Here's a response about why complexity theory may not constrain AI that much: https://www.gwern.net/Complexity%20vs%20AI

It is something "AI alarmists" think about.

What happens if you give the human player a handicap? I wonder if the games are really as close as the commentators say, or if it's just a quirk of the MCST algorithm.

The same entity should not be responsible for both education and certification/credentialing. It creates a serious conflict-of-interest, and a better system would be to separate the entities that grant degrees from entities that sell education. This is especially important if the purchaser of the education is a third-party like a government or a bank, and not the actual student. (i.e., when the education is "free")

Doesn't that possibility just make it easier to put the sheriff on the bottom though?

I think the harder to judge possibility is if RH and crew were locked up rightfully for committing some sort of crime.

Eh, it seems to me the sheriff horribly abused his position of power. Either LJ and RH were imprisoned rightfully, in which case perhaps they shouldn't be just be released, or else the sheriff wrongly imprisoned them. Either way, he exploited his power as sheriff and blackmailed MM for sex.

RH's actions, on the other hand, while certainly misguided, are at least understandable, if he lives in a typical culture of masculinity, monogamy, and Victorian-era purity, where it's embarrassing to be rescued by a woman and shameful for women to have sex. He was probably feeling emasculated after being rescued by in such a fashion, and took out his anger on MM. (Which is bad, of course, but I don't think it's worse than what the sheriff did).

Finally, RH's "abuse" is left ambiguous, as opposed to the sheriff's which is very much explicit.

Interesting, but I'd be surprised if anyone who reads HN chose anything other than some permutation of {{Maid Marion, Little John}, {Robin Hood, The Sheriff}}, which cuts down the answer space by a factor of 6.

My own ranking was MM, LJ, RH, S. Did anyone choose something different from that? Or seriously choose anything different than one of the four permutations above?

Are you able to think in lojban, then?

The idea behind lojban appeals to me, but (according the the times article currently on the front page of HN), there's an active community of Esperanto speakers in NYC.

I'm not really interested in extending the number of people I can communicate with, but if I learn a new language, I would like to be able to actually speak it out loud with other people.

Learning a conlang seems like fun. But should I start with Lojban or Esperanto?

My impression is that Lojban is meant to be as logically unambiguous (in terms of grammar) as possible, while Esperanto seeks to be more of an international/universal language. I think Esperanto is more widely spoken?

Both are apparently somewhat easier to learn than other languages.

So which one?

And what attracts top-notch faculty? Other top-notch faculty and an excellent reputation. You end up with roughly the same positive feedback loop.

Of course science is dangerous. The study of atomic physics very nearly led to nuclear war.

Dangers and ethical issues only arise when science is applied as technology.

I mostly disagree with this sentiment. If you're a nuclear physicist, and you publish a paper about the theoretical aspects of fission chain reactions, and realize that they may allow for good things, like nuclear power, and bad things, like nuclear weapons, you have an obligation to consider the consequences of how your research might be applied by others, do the cost-benefit calculation, and decide to publish, or keep your knowledge a secret[0].

Of course, in the vast majority of cases (GMOs, medical research, human intelligence, etc., knock yourself out) it's probably better to publish - the benefits far outweigh any costs. But I see no reason why this has to be universally, fundamentally true - maybe there are secrets of the universe too dangerous for publication - knowledge which, if made public, would predictably lead to great destruction.

[0] http://arxiv.org/html/physics/0207094 Leo Slizard is usually credited as the first to come up with the idea of weaponizing nuclear fission. Slizard was in favor of secrecy, Fermi was in favor of publication, and for a while, they kept some of their discoveries a secret. In the hindsight of history, this was probably not necessary, but at least they weighed the consequences and had the debate.