Posts
Comments
Backup power supplies are going to be good for a given amount of time. How do you plan to regulate how much of an outage should be possible to buffer with backup energy generation?
In a scenario where multiple AIs compete for power the AIs who makes fast decisions without checking back with humans have an advantage in the power competition and are going to get more power over time.
Additionally, AGI differ fundamentally from humans because the can spin up multiple copies of themselves when they get more resources while human beings can't similarly scale their power when they have access to more food.
The best human hacker can't run a cyber war alone but if he could spin of 100,000 copies of themselves he could find enough 0 days to hack into all important computer systems.
And I think I'm pretty good at not straw-manning people on these issues; I'm used to drawing pretty fine distinctions between pretty out-there ontological and epistemological views.
To have a conversation where you can draw fine distinctions both parties of the discussion need to be willing to have that conversation. You might have approached the issue as a mistake theorist while they were more in the conflict theorist mindset and didn't want to yield any possible ground.
I remember in a book about applied ontology by a author who saw himself as a realist while the statement that "homo sapiens is a species" was seen as a statement about reality while "Alice sake" in the sense of "He did X for Alice's sake" is not something that really exists. ( I think it was in Applied Ontology: An Introduction by Katherine Munn and Barry Smith)
In contrast to that non-realist ontologists consider the idea of a species to be constructed and not one about reality. What's real about the statement "homo sapiens is a species" is not something that's directly made up of atoms and physics but the realist position by applied ontologists like Barry Smith is still that it's real.
Elon Musks Starship might bring us a new x-risk.
Dropping a tungsten rod that weights around 12,000 kg from orbit has a similar destruction potential as nuclear weapons.
At present lunch prices bringing a tungsten rod that's weighted 12,000 kg to orbit has a extreme cost for the defense industry that was labeled to be around $230 million a rod.
On the other hand, Starship is designed to be able to carry 100 tons with equals 8 rots to space in a single flight and given that Elon talked about being able to launch starship 3 times per day with a cost that would allow transporting humans from one place of the earth to another the launch cost might be less then a million.
I found tungsten prices to be around 25$/kilo for simple products, which suggest a million dollar might be a valid price for one of the rods.
When the rods are dropped they hit within 15 minutes which means that an attacked country has to react faster then towards nuclear weapons.
Having the weapons installed in a satellite creates the additional problem that there's no human in the loop who makes the decision to launch. Any person who succeeds in hacking a satellite with tungsten rods can deploy them.
Very few people rent bare plots of land. What's rented (and what was referred to in the OP as rent)lot of land itself, on one hand, and the infrastructure and buldings built on top of it, on the
It seems like you are withdrawing to the motte. Very, few people rent bare plots of land. What's rented (and what was referred to in the OP as rent) is rather floor space in apartments.
The first city that came to my mind was Tokyo. I looked it up and the urban land prices decreased from 885k in 2009 yen per square meter to 758k yen per square meter in 2012. Prices also fell in Tokyo after the bust in 1992.
In completely unrelated matters, Tokyo is also the city without zoning laws...
Fixed it.
At least with the present immigration laws there's a limit of how many people could move to a city. The bay area has a population density of 335/km2 while Manila has one of 20,785/km2. If you increase the density of the bay area to that of Manila you could house every US citizen in it.
If you grant that some people actually want to live in other cities like Boston, there's enough city space for people to live in even without having to go to the density of a place like Manila.
You just have to be serious about cities needing much more density and you can fit everybody in them.
It seems hard to be certain what the true probability happens to be that global warming increases the frequency of floods. On the other hand it's easy to be certain about the probability that the latest IPCC report gives to that event.
In a similar sense you can ask about "How many percent of Muslims in the UK believe according to the GSS that homosexuality should be punishable by death?"
When talking about deaths due to Chernobyl you can argue whether you should or shouldn't count those who got depressive and committed suicide afterwards but you get a clear number from the UNSCEAR assessments of the Chernobyl accident.
If you word your questions to say "What does authoritative source X say about Y", you can have certain answers.
There's prior art.
Stefan Schubert created along with Spencer Greenberg created along with https://programs.clearerthinking.org/political_bias_test.html (https://www.lesswrong.com/posts/3jeRSuLopjjFwFpoG/political-debiasing-and-the-political-bias-test. There was a LessWrong discussion beforehand that informed the questions on that test.
https://www.lesswrong.com/posts/3jeRSuLopjjFwFpoG/political-debiasing-and-the-political-bias-test is an article about his test.
He wrote a guide for creating new tests:
https://docs.google.com/document/d/1OIiIVrNwK6ftRZqLfEmhwXHj0tx-Uhg_N2kfBEFpuP0/edit
You admitted a mistake but it wasn't the mistake for which I was criticizing you. I don't have a problem with people misremembering numbers. This prompted me to explain my criticism.
You lost your meeting room because you were discussing (what I assume to be) politically sensitive topics.
No, I lost it because it was a LessWrong meetup and there are such discussions on LessWrong (and our meetup.com page says SSC/LW, so SSC association was also a problem). The problem was not that the topics might be discussed on the meetup with was more applied rationality focused.
The problem was one of association, not one of meetup content. We could have held the meetup if we wouldn't link from any LW or SSC branded page and called it 'rationality meetup'.
So now I'm left wondering, if not here, then where?
The 'Darwinian Gender Studies' facebook group seems one place worth mentioning. TheMotte was founded to have a place where discussion could be happen with less collateral damage.
There might still be a risk for any insider to participate in them with their public identity attached. Private discussions behind closed doors would be less risky.
Exactly 50-50 would be very surprising result for a meta-study. "50% heritable" has an exactness that "around half heritable" doesn't have.
Treating both of those the same way is what I would expect from people who don't respect actual numbers.
We lost a room in which we held LW meetups in Berlin because LW discusses topics that shouldn't be discussed. The discussion in itself is 'unsafe' regardless of how you discuss or what conclusions are reached.
That's norms for using a meeting room. When it comes to norms that the gender studies community expects there own members to follow, a person who has a reputational stake in the community has a lot more to lose from violating norms in that way.
This isn't even a question of the academic quality of their discourse. a/atheism doesn't attack people in a way that destroys careers and isn't dangerous to anyone. This is different here. I wouldn't want a lone reasonable person in the gender studies field to lose their social capital and/or career for associating with this place.
The standard way LW historically handled politics is by discouraging it's discussion. SSC did things differently and payed a price for it.
That's all separate from the actual quality of the academic discourse but it matters. As far as the discourse goes https://quillette.com/2019/09/17/i-basically-just-made-it-up-confessions-of-a-social-constructionist/ is an article by an insider where he reflects on the low standards he used over the decades.
IQ is around 50% heritable, the other 50% also matters, though.
This sounds like it's written by a person who's not quite clear what X percent heritable means. Apart from that making up numbers like this for rhetorical purposes and treating them as if they are factual is bad form.
The right answer to the nature vs. nature debate isn't it's 50-50 but: That's a bad question and a bad frame for understanding reality.
Instead of debating nature vs. nature one should look at the empirical findings we have and build up a view on the world based on them.
Any reasonable scholar who's in gender studies faces a high reputational risk if they would debate on LessWrong in a reasonable way about their field. Any field that has dogma's that aren't allowed to be publicly debated has a problem with the kind of open discussion we are having here.
Again, I'm not super confident in this and I think there is a decent chance that this will wind up being pointless but it still seems worth spending a little time investigating.
The question is not just whether it's pointless but about whether it's potentially harmful.
In multiple LessWrong surveys biorisk was rank has a more probable existential risk then AGI. At the same time there's very little written on LessWrong about biorisk. If we could recruit people into our community who could represent that topic well, I think it would be very valuable.
It seems like this particular video basically says:
- "Snoring is really bad and side-sleepers snore less." It seems that actually monitoring whether one snores via an app and them optimizing based on the answer is
- Hunter gatherers sleep mostly on the side. Other great apes also sleep more on the side.
- An argument about the waste clearing the body working better on the side (I don't know how strong his case is on that point)
In total it says there's no strong evidence but the evidence they reviewed point to side sleeping being better
Depending on standards of what counts as sexual misconduct, men in powerful position quite often engage in it.
He basically engaged in extramarital affairs and wasn't always fully truthful about them with his wife. He also seemed to have paid money to sex workers which is also an activity in which many people engage.
As far as people we know of in that kind of position of social power, this doesn't seem untypical to me.
It sounds like you are suggesting that a person can simply decide not to have opinions. I find that doubtful as most people have little control over what goes on in their head.
When it comes to deciding whether or not your approach has merit, it would be necessary to analyze the actual likely effects from following it and what it means to follow it for a human.
It seems like the official story as you for example find it on Wikipedia says that the system detected five ICBMs.
You can't decide whether or not someone made a reasoning error by engaging in an action without understanding the goals that the person has for engaging in that action. The goals are part of his mind and the fallacy in which you engage by assuming you know the goals is mind reading.
You seem to treat the situation like you are the first person whom the interviewer asked those questions. In most cases that's not true and the fact that your friend faced the same questions is evidence for it not being true. You should expect that the interviewer knows the kind of answers the average interviewee gives to the question.
The situation isn't similar at all. In the case of Henry Ford the question is whether there's coherent concept of ignorance under which Ford as a single individual can shown to be ignorant. It's also not a situation where people get ranked against each other based on their ability to deal with the same questions.
That said, the case of Henry Ford was a win for everybody involved. Ford got publicity, the newspapers had a good story to sell and had a judgement of 6 cents against them and didn't have to pay Ford the 1 million in libel that Ford was asking for.
It seems to me that your reasoning here mixes egocentric thinking, black-and-white thinking along with mind-reading.
Egocentric/Black-and-white: When interviewing people, the interviewer usually doesn't care about individual performance but about how different interviewees compare with each other. If something is a question that's hard for everybody to answer,
Mind-reading: It's possible that the interviewer didn't cared at all about your memory but cared about your ability to give decent answers when you don't know the technical details of what you are asked about. That seems to be a skill that matters to a product manager.
The focus of trying of doubling down on antibiotics seem to be the wrong strategic choice. It's likely that we don't use any antibiotics in a few decades anymore, not because they become ineffective but because antibiotics are blunts untargeted instruments.
Gene sequencing allows us to understand the cost of lost biodiversity among bacteria in the gut better and allow for optimizing cut bacteria.
More importantly gene sequencing allows us to understand the actual bacteria that causes an infection. Antibiotics are currently superior to phage therapy because phage therapy needs to be targeted to the particular bacteria you are dealing with while antibiotics don't. Once we actually sequence all of our infections and thus know what we have to target phage therapy will be superior.
Okay, when you already have a system where there exists only one state approved nationwide textbook this would be easier politically as it is in states like the US or Germany where that isn't the case.
Currently, textbooks are made by private corporations and not by the state. It seems to me like your proposal does change that and would make the state more responsible for the textbook.
If you make the textbook free you don't have different vendors competing anymore into providing a superior product.
There are plenty of ways to donate anonymously already when that's what people want to do that take a lot less effort then changing a name.
Your proposal might be conservative in the political sphere of your country where the people who want socialism back are the conservatives but in most of the Western world using socialist means of centralized planning is considered to be left-wing.
In particular this proposal centralizes power over how the textbooks that are used to educate the young into the hands of the national government and thus is likely to be both opposed by the teachers who want that power to be at a lower level and the companies who make money with selling services into the education market.
Taking on both the teachers unions and all the pro-corporate lobbyists is a political nonstarter. It's the complete opposite of an uncontroversial proposal.
doubling the size of a tax quadruples its social cost, so it’s better to have lots of small taxes rather than a few big taxes
I doubt that's the case. Compliance costs scale roughly linearly with the number of different taxes.
Compliance costs with filing taxes in the US were roughly ~200 billion in 2018 for a tax revenue of ~3300 billion.
The solution you propose also sounds really complicated when people have to optimize the timing of when they make capital gains with times when the risk-free rate is low.
Basically, there was political discussion happening with non-PC views. As far as I remember EY thought that it was bad to have that group associated with the LW brand and wanted to ban people. There was drama and the group was renamed into Brain Debugging Discussion.
One way to prevent this from happening would be to ban political discussion explicitly.
I don't think there's evidence that MOOC failed because people lacked the skills to complete them. It seems like motivation was a much bigger issue and even courses with little skill requirements have high drop-out rates.
I know multiple people who sell self-help material where people would just have to watch a bunch of videos and could watch them without doing any exercises but where still most people who actually pay money fail to watch all the videos.
But as things are, I expect that even if one of your meetup experiments failed, it would give us useful data.
It's one thing to run a meetup experiment. It's another to globally say that everyone should run their meetups in a certain way.
Global coordination needs much more buy-in from other people.
I’m also in the process of creating a Facebook group for attendees of all meetups worldwide.
There used to be a LessWrong Facebook group and now there isn't anymore. Are you aware of what happened? What kind of governance do you want for the new group?
I'm not sure how "official meetups" would be any different by nature of being "official" or even what official means. The idea also seems a bit strange to me because I don't think the LessWrong.com team has any claim of being an official arbiter on the term LessWrong.
LessWrong Germany e.V. is a NGO with a 5-figure yearly budget.
It seems to me that an individual toastmasters club has a lot less license to innovate then our local meetup has. A toastmasters club can't say: "Let's run this marathon together under the logo of our club." but our local meetup would have no problem with running a marathon together as an LessWrong team.
It seems to me like making a top-down decision about what topics people should discuss is a way to remove agency from individual meetups. Our meetups in Berlin also aren't discussions about topics but rather about doing rationality exercises together.
I don't seem why it would be important to have such a social standard on LessWrong. We already have a problem of people on LessWrong being discouraged from writing post because they expect to get criticism that isn't helpful to them.
I rather have social standards on LessWrong that do enforce quality norms in criticism then encourage people to voice criticism without providing arguments.
How about asking the user when they click on "Community Events" to allow the browser to get their location? Then if the user says yes, use that location to show the right events on the left side of the front page.
How much does it help that you can watch what the police are doing when they still have all the guns? Maybe not such an issue in America, but what about Hong Kong?
Public officials in the US get punished very seldom while in China, it's much easier to throw public official into prison.
China does a lot of public opinion management because public opinion matters to powerful people.
I don't think LessWrong is unique in that regard. Wikipedia is strongly focused on it. The StackExchange network also has a lot of content that's intended to be available in the future.
Do you mean there are now 10 left in Leverage proper?
On an IOT sensor that triggers whenever the door is opened.
Science and Sanity which coined the phrase "The map is not the territory" does contain a long discussion on what can be meant by saying a map is accurate and how abstraction works in this context.
I would recommend reading it if you are interested in the nature of maps.
I do think that there are seldom cases where I would tell someone "be more angry". I might however say:
- Don't withdraw yourself from situation that might produce anger.
- Don't suppress your anger
- Have personal boundaries (which can produce anger when they are violated)
Those suggestion might lead to a person feeling angry more often but they have a different focus.
If every study on depression would use it's own metric for depression that's optimal for the specific study it would be hard to learn from the studies and aggregate information from them. It's much better when you have a metric that has consistency.
Consistent measurements allow reacting to how a metric changes over time which is often very useful for evaluating interventions.
A possible fix is getting an e-bike.
Neuroticism is one of the big-five personality trait. Personality traits aren't things to be managed moment to moment.
You generally want to minimize stress and maximize output. Instead of thinking "It would be good to be more neurotic" it would be better to go "I need more goals that I pursue that I care about".
Planting trees for the sake of the environment is not a crazy idea. It's a mainstream idea that's held by many people. You can buy a beer and in the process support protecting space in the rainforest.
hereisonhand spoke about him being able to extend that idea directly into "survival of the amazon rainforest was a significant climate change initiative".
To me that suggests he sees viability as the same thing as it being a good action. It does look to me like reasoning about public interventions without the EA mental models that are needed in the context to reason well.
Thinking in terms of internal parts is a mental model that a good portion of the LW community that's interested in self-improvements techniques can use. You need it for the Internal Double Crux technique that CFAR teaches.
Yet, it's not the only model out there. I personally rather do a version of Leverage's belief reporting that assumes I as a whole either hold a belief or don't then doing parts work if I believe that a specific belief that I can identify is the issue.
As far as abstraction goes, I think it's a key feature for self-introspection. If you are mentally entangled with the part that you are introspecting you won't see it clearly.
A lot of meditation is about reaching a mental state where you can look at your thoughts without being associated with them.
You are right that there are contexts where viability is a useful notion. It just isn't here.
"Viable initiative" is concept that isn't useful. What we care about are "effective strategies". Whether or not planting trees is effective depends on how much it costs and how much other strategies cost.
What does that even mean? Oxygen doesn't get produced. What trees do is that they bind the carbon in CO2 and leave the oxygen molecules in the air.
I think it's wrong to call this a criticism of the rationality community. The people who designed systems like market auctions for spectrum aren't member of the rationalist community.
When I'm thinking about an institution in our movement that advocates knowledge gathering through formal methods without attempts at openness no one come to mind.
You could call GiveWell an organisation that uses formal methods for knowledge gathering but they are also an institution that releases recordings of their board meetings to the public. Actions like that are a costly signal for valuing legibility.
CFAR doesn't teach people to reason with formal systems. That's not what they teach.