brownbat

I've heard this rebuttal a few times, that western values try to suppress emotions, and anger is sometimes good. Eliezer makes this point too. I'm still not sure I fully understand it.

The anger I feel when I'm cut off in traffic is distinct enough from the anger that I feel at oppression or injustice to really deserve its own label. Like, anger1 and anger2.

Since they feel so different, I have slightly different thoughts about each.

I never want to drive with anger1 instead of safety. Driving with anger, or "road rage," is not an interesting value-neutral way to experience the world.

I think that driving anger is a distinct emotion that is always unhealthy and just happens to share the name "angry" with other emotions because of superficial similarities. People will probably say that's just a different context, I think that's overgeneralizing, but even in other contexts I worry about the call to set aside rationality. (Instead of anger1 and anger2 you could call one rage and the other indignation.)

When I'm angry ("angry2") about senseless human suffering, I agree it should move people, people should not just tune out. But that seems like a straw man of the rationalist approach. There too I'd rather channel that feeling into a thought process that maximizes my impact, rather than just attempt to satisfy the emotional call to yell or cry or punch a wall. Those may be important therapeutic responses, but if you can feel satisfaction by just getting to work to try to correct the injustice, that would be better.

I'm rarely physically threatened, but my system still occasionally floods me with fight or flight adrenaline. What's so wrong about saying that's a quirk of evolution, and thoughtful responses are going to leave me better off, absent some weird fringe scenarios?

I've heard the opposite perspective from people smarter and calmer than me though, so maybe I'm wrong and it just hasn't clicked for me yet.

We are naturally far better at identifying what's plausible instead of sorting out what is likely, and we're better at giving kudos and recognition for the former than the latter even though it's easier. That makes some people hyperdevelop "theorizing" without working on "validation."

We are dangerously good at making up stories to explain data. Talk to anyone about a spurious correlation[0] and they can provide you a just-so story that explains it as a real correlation in under a minute.

Provide anyone with two competing theories that each account for some but not all of the available data, then ask them to rank the two based on likelihood. You'll make people visibly upset. (Our ability to invent new theories to fit any circumstance might partly explain why, for a while, explorers forgot how to cure scurvy.[1])

And it's easier to "look smart" the first way. No one can immediately determine whether you're good at the second unless you make a long series of predictions evaluated over time while constantly gathering more data. Everyone can see if you've connected a few dots in a clever and unexpected way. Talking heads are regarded as experts for going on shows and making vague predictions based on a few recent events, but are never called back to scrutinize their predictive track record.[2]

Conspiracy theorists are generally just arguing something is plausible, and ignoring the messy step about what's more likely, with the implication you should just leap from plausible immediately to unwavering conviction.

Some of this is just, well, one of these things is hard and the other is easy. But I'm confident both can be learned to some extent, based on personal experience of getting a better sense for these over time.

So to the extent this is "fixable" at all,[3] maybe greater education about meta-cognition (esp. cognitive biases), the history of analytic philosophy, a humility about what one knows, and/or Bayesian reasoning earlier in education might help.

Sidenote, this itself is arguably just another just-so theory too. Maybe I'm wrong about it all. :)

[0] e.g., http://tylervigen.com/spurious-correlations

[1] https://idlewords.com/2010/03/scott_and_scurvy.htm

[2] This is Phillip Tetlock's whole area of research.

[3] To be fair, maybe we've already "fixed" this pretty well, we tend to attribute far fewer things to witchcraft than a few hundred years ago, maybe normal everyday education and expansion of literacy is enough and there will always be a handful of outliers and trolls. See also: https://slatestarcodex.com/2013/04/12/noisy-poll-results-and...

> Voting in party primaries and advocating for change within parties, given the state of affairs in US politics, is a better bet for effective change than maintaining a stance of moral purity.

I don't see why this would necessarily be true. Unaligned votes decide elections, while deeply partisan groups can generally be assumed safe and disregarded. Turnout boosting is a thing, but not generally through policy concessions. Turnout operations are much more about rallying people who already agree with you.

I might be wrong, but my strong hunch is that if you have an hour to spend, volunteering at a party call center is going to pale in comparison to just writing a carefully worded letter to your representative on a specific issue you care about.

Because it's just emotive, "boo mainstream!" which doesn't generally admit a detailed rebuttal. But here goes -- Wagner's main argument takes an action, ties symbolism to it, then denounces that action because of the author's wholly invented symbolism.

But anything can symbolize anything with the right narrative attached. It's just word games to criticize people when you don't like their personal tastes.[0]

Even if weren't for the flawed reasoning, the sentiment is pretty important here. If someone started going around dismissing other commenters as "normies" in these threads, we wouldn't demand careful rebuttals and carefully pull out their good points. We'd just flag them and move on. But that's the whole climax and title of this article! Why is it suddenly cool here? Because she's leveling insults at people we might not like?

Is "don't insult people, unless you happen to not like them," really the threshold?

[0] She does the same thing on McMansion Hell, which while sometimes amusing, has basically morphed into this whole attempt to cast people as deeply immoral because they wanted a skylight where she wouldn't have put it. And people love her for it, and that's great, but people love cynically piling on and judging other groups of people for the most trivial reasons. Doesn't make it nice.

> At the last conference I attended, I asked most people I met two questions:

> 1. Do you know of any companies that aren't highly dysfunctional?

> 2. Do you know of any particular teams that are great and are hiring?

> Not one single person told me that their company meets the criteria in (1).

Reminded me a lot of Ribbonfarm on the Gervais Principle:

"Hugh MacLeod’s cartoon is a pitch-perfect symbol of an unorthodox school of management based on the axiom that organizations don’t suffer pathologies; they are intrinsically pathological constructs."

https://www.ribbonfarm.com/2009/10/07/the-gervais-principle-...

Yeah, passages like this:

> A side effect of the fetishization of the industrial is that it aestheticizes the backdrop of labor within capitalism, in many ways erasing labor’s histories of toil and struggle.

"Why, of course a white kitchen signals you as an enemy of the people. It's so obviously counterrevolutionary!"

I imagine stern, angular posters, "Mind your palette, comrade!"

The piece isn't just Marxist, it's Soviet.

But also... schizophrenic? You can't idolize the humble worker while hating "normies." The top 1% alone don't make something ubiquitous. The eccentric preferences of billionaires aren't driving the aesthetics that take over Target and starter homes or get featured on reality TV.

This is incredible!

I noticed that if I try to change stations in Seoul (track forward or track backwards) that part of the interface freezes with a "loading..." message.

Tiny bug, probably easy fix.

Such a great experience overall. If you plan to keep building it up, would love even more cities, even "boring" ones, if adding them could be pretty well automated. Synching the time of the video to the time in that place would be an amazing feature, but would probably involve a lot of tagging, and not sure if there's that much content out there. Keyboard shortcuts, including one to hop to a new random city would be nice. (I guess ctrl+r technically does that already...).

EDIT: There's a whole genre of 4k city flyover videos with drones on YouTube too. I know it's off concept, but that might be worth remixing in a similar way (assuming you could give credit and get traffic back to the original producers).

Great stuff.

To echo your point, there was an Aspen festival panel a few years ago where a network news rep/owner was refreshingly honest.

He said everyone thinks they want the news to be deeply insightful and informative, but the people who would consume that product tend to go out nights and do interesting things like attend ballets or go to museums, or go to things like the aspen festival. Or at least they are just reading a few articles online.

They make news for the sort of people who just binge television news constantly. These people don't have a custom rss feed to make that efficient, and that is a special type of person.

"We run the market test against PBS every night."

If PBS started hitting numbers, every other station would jump to that model immediately.

Yeah. What's frustrating is I feel like a lot of ports have long load times just coded into them. Sometimes it barely matters if you run them on an NVMe or even a RAMDisk.

I'm sure the cost of speeding up loading isn't worth it after you've written it with one set of expectations in mind.

I just hope an unintended side effect here is that load times for big studio PC games speed up too.

> "A 100% open and free journal cannot achieve selectivity without having some judgement and bias applied."

Establishing reputation is the central challenge for a lot of the internet. Sorting spam from mail, sorting useful search results from SEO, sorting legit programs from malware on app stores.

"Let's just have a small handful of people manually review everything" is not a terrible first approach! It is the naive solution, and will work if you don't have to scale. It even worked for search for a couple years.

And you might argue that it's ok for journals to keep doing that because they don't have to scale. They don't have to review, rate, and publish everything good. They can have a very, very tiny output and it's ok.

But there is some cost to rate limiting scientific output.

So I'm surprised there hasn't at least been a good competitor incorporating what we've learned from other domains. It wouldn't be the same, but at least trying to use some things like citation counts and reader behavior for an initial guess at what deserves review.

All the arguments that "we need a small group of professionals curating these" lose a little weight in a replication crisis.

If you really wanted to try this, you might want to go after low hanging fruit. Someone should make a nutritional science journal, using purely algorithmic data to score proposals. Not much to lose there.

https://io9.gizmodo.com/i-fooled-millions-into-thinking-choc...

This led me down a rabbit hole into the sterile insect technique, where large numbers of sterile insects are released to reduce the population.

Looks like Alphabet's Verily has been testing it on mosquitoes in Fresno: https://blog.debug.com/2020/04/nature-biotechnology-features...

Schuyler Towne has (several?) great talks on this, going through the history of physical security and our assumptions about perfect security vs. social and cultural symbols.

https://www.youtube.com/watch?v=jqjacHSTd48

I'm surprised those who have survived covid haven't found a way to self-identify, like green wristbands or something that say "immune" or "survivor," and wear those in place of masks.

Not recommending this, it would still create some of the same problems, but I'm surprised it just hasn't naturally emerged from individual incentives.

> When I used Google Play Music a couple years ago there was a feature that was broken for YEARS

Not sure if this is yours, but mine was that it would often swap out my tracks for radio edits, even though I uploaded the original. I'm an adult, I bought this music, I don't need the kids' version. There's a built in feature you can use, "fix incorrect match," that would fix this only about half the time.

Oh right, it also capped the max size of playlists to 1000 songs for some reason.

Now it sounds like YouTube Music won't even let me listen while switching to another app, like a workout app, and won't let you turn off the screen without the music going out. Hard to believe.

Looking around, I'll be moving to musicolet I guess. Sounds like AIMP and BlackPlayer get good reviews too, but I really love the idea of an app that actually tries to limit the permissions it requests from you.

https://www.androidauthority.com/best-music-players-android-...

Aza Raskin's Ubiquity was such a clear model of how to build voice interfaces the right way, and it wasn't even a voice interface. It was a bit of a launcher that tied APIs together on the web.

Let users create and share small commands. Create a simple natural language for commands that are easy to program, extend, and remember, and narrow the scope of inputs the voice engine has to deal with.

It was so beautiful and effective and just light years ahead of what we're getting.

Microsoft gets a special mention for lost potential here. Their voice system in Windows could be a way to navigate the layered menus of the OS, but it is mostly focused on answering general queries. Voice is a great replacement for the program launcher, except it's not customizable, but that's about the extent of how much you can control the system with it. Let me do anything buried in the control panel, show me everything you know about a process when I ask, solve that first, then worry later about telling me how big the moon is. You make an OS, don't forget what that is.

A group of us once tried to rank cities in Europe by population only to realize that most of them are effectively incomparable.

Cities sometimes have clear legal boundaries that feel irrelevant to the question, like the City of London, but more generally have metro areas that sprawl well into an ambiguously defined countryside. There's rarely a "this block is city, the next block over is clearly not" situation, so the number of people you include ends up being pretty arbitrary.

Yeah, if you relay content at scale, your first preference is to do nothing for moderation. After a little taste of that, people get upset and complain, which is just a little inconvenient at first.

It becomes progressively more inconvenient once some of those people with complaints can subpoena you or fine you millions of dollars or exclude you from large markets.

So your next best option is hamfisted moderation.

Tuning things costs money and time,[0] it's generally easier to just shoot the moon on either false negatives or false positives.

[0] and given conflicting preferences from different powerful groups, is arguably impossible anyway

Or, "I tend to go with philosophers from Voltaire to Mill to Popper who say the only solution is to let everybody have their say and then try to figure it out in the marketplace of ideas. But none of those luminaries had to deal with online comment sections."

https://slatestarcodex.com/2019/02/22/rip-culture-war-thread...

EDIT: I should add, Podcast Addict is my go-to. I get how this can happen, but this is all really annoying and I feel for Xavier Guillemane and all the other users. Here's hoping the massive press blitz on this gets a quick reversal in this case, and Google finds a way to prevent it for apps with less visibility. Maybe we can enforce tuning by getting as upset as possible at the false positives.

I remember a post here a while back from a guy running Bitcoin mining in his dorm room. One day he realized he could offer his spare cycles to grad students with high computing workloads and undercut cloud computing prices while increasing what he made far over just mining.

It's very weird that we haven't fully arbitraged $/instruction to a single (low) price yet (or storage/hosting, whichever).

If only there were an Uber for unused cycles or storage... let everyone turn their unused capacity into mini AWSs with a common interface and safety and reliability guarantees.

Maybe there are too many barriers, like security.

I've been there. Some article or art really lands with me, falls completely flat with friends or when I post it.

I've also experienced the opposite--a focus of this very article actually--creating something that catches on much more than I expected.

It's surreal being able to create something that people love and not being able to exactly know why that one bubbled up in particular.

Feeling a profound connection to some work and not having others see what you see is the brother of that confusion.

If you want a surprisingly deep meditation on popularity in art, Chris Rock's Top Five has some interesting dialogues on it. And Chuck Klosterman's "But What if We're Wrong?" is a good book basically about how unfathomable popularity is.

Trains are far more efficient, less subsidized, safer, less polluting...

https://grist.org/article/freight-trains-19th-century-techno...

Even if you electrify trucks the tire and brake wear and road degradation contribute an enormous amount of PM 2.5.

Sure you'll need trucks for the last mile but every ton of freight we can move to rails or waterways is a victory for humanity.

> the IE regulatory situation in Korea

I found this from 2013:

https://www.zdnet.com/article/south-koreans-use-internet-exp...

Anyone have an update to that story? I know in Korea they use a ton of apps to order things all the time now. Maybe this law was phased out, or just didn't apply to phones?

Not to take away from your broader point. Laws are just such a terrible development model. Everything's tested live and you can't easily revert anything.

I should have said "not uncommon" rather than "ubiquitous," but it appears it already is blocking malaria development in a small portion of mosquitoes.

https://www.nature.com/articles/s41467-020-16121-y

Uncanny. I was also going to recommend that, maybe "ruthless prioritization," to emphasize that everyone is constantly busy and that these choices are sometimes unoptimizably hard.

It would also draw a bigger distinction between the "strategic" decisions he describes at bigcos (which also involve opportunity costs!)

Terminology aside, I did think he made a good point.

Here's a really comprehensive overview of Microsporidia -- https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5613672/

Still working through this and the original paper, but while it can cause encephalitis in humans, it seems to be ubiquitous already anyway, and we don't seem to typically get it from insects. Setting aside the fact that you could swap malaria for almost any other infectious disease and humanity would be massively better off.

(Caveat: IANA Microbiologist and welcome corrections.)

Original study in Nature: https://www.nature.com/articles/s41467-020-16121-y

> All of us should be regularly lobbying our lawmakers to make student loans dischargeable by bankruptcy,

I thought so too. This always made me so angry. Last time I posted a rant about that here someone urged me to read the REPAYE Act, and it completely changed my mind.

Under REPAYE, students get income based repayment, even down to $0 payments if you're anywhere near poverty level. After a certain number of years of repayment, the government just pays off your entire balance.

This solves the same problems bankruptcy is designed to solve without ruining people's credit. It also prevents doctors from declaring bankruptcy right after medschool -- I'm honestly unsure whether that was a real or just anecdotal problem, but income based repayment for a fixed term seems like a better way to let people escape debt while still expecting them to pay what they can.

The course organizer believed in intentional dehydration, so there was no water until the 11 mile mark, with some trainers instead using alcohol when runners became thirsty. It was 90 degrees in the shade.

The winner was periodically dosed with strychnine mixed in brandy, and suffered hallucinations for the last 10 miles of the race, was carried over the finish line, and would have died if doctors didn't immediately treat him.

Another competitor hadn't eaten in 40 hours, so he stopped at an orchard and ate some apples that turned out to be rotten and gave him incredible stomach pains, so he had to lie down.

Another competitor almost died from the dust clouds kicked up by the cars driving ahead and behind the participants.

Another competitor was chased off the race by aggressive dogs.

Just so many crazy things all at once.

There's a fantastic episode from The Memory Palace about the race:

https://thememorypalace.us/2010/02/episode-26-citius-altius-...

Since ammonia and LOHCs face some overlapping constraints I don't expect an order of magnitude difference in cost. But you're right, it would have been more useful with an explicit comparison.

If you are seeing stuff that suggests ammonia will get very close to parity with oil, that's really fascinating, I hope it pans out that way. The IAE is more optimistic than I am about ammonia transport, while still listing some significant hurdles:

https://www.sciencedirect.com/science/article/pii/S036031991...

I think it's a good report, and you could read it to be more convinced of either of our positions.

Sorry we're talking past each other. You did actually change my mind a bit, if it's any consolation. Agree it's too hasty to say electrons are always better than atoms, though I'm still optimistic about more supergrids. My main remaining concern is just that, as you change your power mix away from oil to generate more ammonia, or electrolytic hydrogen, or whatever, the case for shipping fuel erodes significantly, because you can just move generation closer instead.

Sorry we couldn't get to ground on this one. Thanks for your points all the same.

You're right to point out that tankers are phenomenally efficient, but that's _for oil_. Just as LNG is much more expensive to ship than oil, because of the increased handling complexity, any shipping of hydrogen is going to lose that advantage completely to electrification.[0]

> there will be (very long) distances that you'll want to ship energy

This is probably the real crux driving the difference in how we see this.

The main reason we have to ship hydrocarbons long distances is because those resources are so concentrated. Wind and solar aren't perfectly distributed, but they are so much more distributed that local generation would be much more efficient than transporting anything much beyond what current HVDC projects can easily handle.

So, sure, we're not going to stretch a cable for 12,000 miles. But we're also not going to build a solar farm in Saudi Arabia to generate hydrogen to ship to Canberra. We're just going to put a solar farm in the outback instead.

[0] "Relative costs of transporting electrical and chemical energy" by Saadi, Lewis, and MacFarland, studies this very question.

Even in that report, you could quibble with a lot of their assumptions that I think bias it against HVDC, project lifespans, ignoring loading and unloading costs for tankers, etc. But it's not a crazy starting place, and HVDC still beats hydrogen on $ per mile per Joule. My takeaway from it was that if we're stuck with a carbon economy, sure, we'll keep using tankers. If we're talking about going carbon neutral, the difficulties of working with hydrogen relative to oil and LNG will probably tip us back towards electrification and local generation, except in the most highly specialized of situations (fueling Antarctica or something).

The report was a good reminder why we have an oil-based economy though. It is so obnoxiously practical and well suited to everything we want to use it for. Ditching carbon will be hard.

You're right that there's a reason we constantly transport so much liquid fuel around the world.

But the Europe and North Africa connection is already a relatively serious proposal precisely because of the efficiency gains. It's definitely an ambitious project, but more from the politics than from the technology. Electrons are just easier than atoms, even underwater. HVDC has made long distances realistic for cables.

China's Anhui to Xinjiang HVDC link will be around 2000 miles. There's a planned HVDC submarine cable from Lincolnshire to Denmark that will be 472 miles.

So for Australia to China, we might be able to pull off Darwin to Jakarta, then Jakarta to Hong Kong within 10-20 years, if you really wanted to. Would break some records, but it's not crazy impractical.

Though with the most efficient transfers you're taking some percentage hit per mile. Closer generation is generally going to be smarter, especially if you're running cables through what would be really viable offshore wind fields.

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

https://www.power-technology.com/features/chinas-mega-transm...