LESSWRONG
LW

by James_Miller
1 min read
93

14

This thread is intended to provide a space for 'crazy' ideas. Ideas that spontaneously come to mind (and feel great), ideas you long wanted to tell but never found the place and time for and also for ideas you think should be obvious and simple - but nobody ever mentions them. 

Rules for this thread:

  1. Each crazy idea goes into its own top level comment and may be commented there.
  2. Voting should be based primarily on how original the idea is.
  3. Meta discussion of the thread should go to the top level comment intended for that purpose.
Open Threads
Personal Blog

14

Crazy Ideas Thread
New Comment
93 comments, sorted by
top scoring
Click to highlight new comments since:
Some comments are truncated due to high volume. (⌘F to expand all)Change truncation settings
[-]James_Miller150

Someone should create a free speech Twitter that doesn't censor anything protected by the U.S. 1st amendment.

Reply
8buybuydandavis
Decentralized twitter where nothing is blocked and everyone subscribes by the rules they choose, including delegation of subscription rules.
6knb
I think there is a strong case for breaking up Facebook and Twitter as telecom monopolies. They would be forced to adopt open standards, so anyone could send information to their users, and other companies would be able to create their own clients to send info to facebook/twitter users and vice versa.
0ChristianKl
Quite a while anybody could send information to Facebook users because Facebook implemented the open standard of Email. I think that feature was mostly used by spammers. I don't want everybody to send me messages in an unfiltered way. As far as breaking up Facebook, I don't see a reason why they should have Instagram and WhatsApp but the core Facebook service can't be easily broken up.
1knb
Sure, I could send emails to your facebook account, but if I wanted to see any of your social media content, I would have to start a Facebook account and access it via Facebook's walled garden. If I want to use Google+ and you use Facebook.... It's as though you had to use a Verizon phone to have a conversation with other Verizon users, and you couldn't use your Verizon phone to contact people who use AT&T. The outcome is inevitably a monopoly due to Metcalf's law. Your social network client could still have filters, but the filter would be something you control, and it wouldn't be as arbitrary as "you may only friend-request other facebook users, and only other facebook users may friend-request you." Start with an open standard for friend requests; i.e. Google+ must accept friend requests from Facebook and vice versa. Any new startup would be able to create their own social networking client, capable of sending, accepting, and displaying friend requests, media shares, private messages, wall posts, etc. This would create a much better, more competitive system with vastly more consumer surplus.
4garabik
As a rather illustrative case study look at the history of XMPP at Facebook and Google (Talk). Facebook messaging used XMPP until 2015, but it was not federated - but at least you could use client of your own choice. Then they switched to a proprietary API. Google talk used federated XMPP for years, but then dropped server-to-server encryption, effectively cutting off majority of servers, and then dropped Talk in favour of proprietary Hangouts altogether. So the trend is just the opposite - if the player grows big enough and the community becomes self sustained, they will start walling the garden.
-1ChristianKl
Google+ doesn't have a concept of "friend request" or friend. Forcing them by law to do so, reduces the freedom of companies to innovate.
2root
https://status.fsf.org/fsf Good enough?
1Manfred
Name: Fritter.
1username2
Most likely it would contain a lot of spam and most advertisers wouldn't buy ads there in order not to be confused with spammers.
1hg00
Lower-hanging fruit in terms of making money, I think, would be to create an explicitly pro-meritocracy Github competitor. You could call it "MeritHub" as a reference to the "United Meritocracy of Github" rug that Github famously removed when pressured. Startup ideas like these market themselves. It wouldn't be much work: use Gitlab as a starting point, make it available for open source projects that are leery of hosting their project on Github's servers, and charge money to commercial entites who want to host their code with a company that isn't controlled by leftists. Server costs might get expensive, though, if you had a high ratio of open source projects to pay projects, or you had to survive sustained DDOS attacks. (I actually disagree with your Twitter idea--Twitter is terrible. Memetically speaking, it's the equivalent of ingesting 10000 petri dishes full of bacteria chosen randomly from labs around the world just to see what it does to your gut flora and immune system.)
0ChristianKl
Which kind of companies do you expect to care about his enough to pay money?
-1hg00
Github already has paying customers who are looking for private code hosting solutions. (They host public repositories for free, but it's not wise to keep your code public if you're trying to make a profit.) So, sell to the same customers, but to the conservative-leaning ones. You could even go dirty and write a blog post compiling Github-related SJ controversies, then suggest that they aren't a reliable place to host your code (intimating that if your company became the focus of an SJ controversy, and SJ advocates pressured Github to make your code public, they might do it. Who knows if this is actually a risk or not, but it couldn't hurt to plant it in peoples' minds.) NRxhub.com - a place to host code during this century and the next.
0ChristianKl
When it comes to trust, it's easier for companies to trust big companies like Github than it is to trust a random person who decides to provide code hosting. Even if someone buys that argument, hosting with Atlassians Bit Bucket would make more sense then to host it with a random place. Businesses can be attacked for hosting their code at NRxhub.com and suffer PR damage. I could imagine a trustless solution that works by doing decent crypto to draw an audience but wouldn't expect a simple solution to do so.
0ChristianKl
That might be a perfect usecase for Urbit.
-1fubarobfusco
Someone should pay to install and maintain a printing press and supply of ink and paper, installed in the public square, for all comers to print pamphlets and disseminate their views, ads, rants, wedding invitations, conspiracy allegations, and so on. Surely this would be an excellent and effective contribution to public discourse... and if not, to the wage of the cleaner who sweeps up litter.
1root
Too easily exploitable. It was common in the fax era to waste a lot of ink (paper too but you can't waste more than one paper per paper) by sending a completely black document. There's probably more sophisticated ways to attack such a system, but don't ask me. Go read Bruce Schneier.
-1[anonymous]
Seems like a great idea for a sci-fi or a fantasy story, or even a detective one... Like, there is a small community in a remote village, a conflict of interests, and the possibility of communication either in the daytime, or by leaving anonymous messages sent by night...:) something like SSC's 'It was you who made my blue eyes blue', but with LESS VIOLENCE.
[-]James_Miller150

GiveWell should have book/movie/podcast/video game awards (like the Hugo award) open to voting for anyone who has given money through GiveWell in the past year.

Reply
2Jiro
Because the Hugo award has gone really well.
3James_Miller
You only fight over things that are valuable.
0Lumifer
Like being an alpha and curb-stomping the betas?
-1James_Miller
The fight here is against other alphas and the prize is high status women. The curb-stomping is a signal.
-1Lumifer
Yes, but the question is whether GiveWell should set up an arena for such games :-P
0James_Miller
Yes, if it can sell tickets.
0Lumifer
I suspect naked midget mud wrestling can sell more tickets :-) Or you can set up fights between alphas where the prize is high-status men...
0James_Miller
This just might harm GiveWell's brand name.
0hairyfigment
Not so (or poorly defined); if you want to hurt someone, you can fight over things that would have been valuable to them if you hadn't fought.
0Evan_Gaensbauer
What makes donors to Givewell-recommended charities special?
1James_Miller
The goal would be to raise money for Givewell. They are special because they likely practice effective altruism.
0Evan_Gaensbauer
Oh, so gaining the right to vote in the awards would incentivize to donate through Givewell? Sounds good. What ideas would you have in mind to make these awards attractive, or prestigious? I mean, why would people care? If they're the only voter choice awards in a given niche, that might make them attractive. The unique spin of being a donor getting someone the right to vote might also entice those looking to have fun while doing good. In that case, we might want to set a floor-level to donate to earn a vote, so the system isn't gamed by a bunch of voters who only donated a cent or whatever. If all that makes sense, what that would leave is finding a good way to advertise the awards.
2James_Miller
A floor might be a good idea, although GiveWell would probably welcome having lots of people give even $1 as it would at least introduce themselves to the organization. You are right that publicity would be the key for this to work.
0Evan_Gaensbauer
Yeah, I think the floor should be relatively low as well, just as a technicality. Like, $0.05 USD, or something, even. Really, I was just thinking it would ruin the vote for future awards if the system was ever gamed by a bunch of trolls, as has happened in the past with TIME's Person of the Year, or with Boaty McBoatface. However, that might not be at all likely, or everyone might be good-humoured enough to not mind anyway. How serious would you be about actually making this happen? It doesn't seem actually-crazy, and seems only 'crazy' in the sense that the time spent doing it poses a high opportunity cost in terms of a member of the EA community who would do this doing something else valuable with their time. Asking because I could be that person, and/or could find one or more person(s) who could do it instead/also.
2James_Miller
"How serious would you be about actually making this happen?" My plans don't go beyond this post, although if someone at GiveWell asked, I would be happy to help out.. It would be fantastic if you wanted to pursue this or find others who would.
[-]Daniel_Burfoot120

Large scale heat management: controlling or influencing temperature flows on a geographic (regional or global) scale. Heat management is one of the deep fundamental problems in life and engineering, but humans have never tried to do anything smarter or more ambitious in this area than standard HVAC stuff.

Humans like moderate temperatures, say 55-75 F, but we spend quite a lot of our time in discomfort or even pain because the actual temperature is outside this range. But the problem isn't that heat (or cold) is in short supply, it's just distributed unevenly. This fact hit home for me when I was riding in an Uber because terrible winter weather knocked out Boston's subway system, and the driver told me she had just returned from a trip to Brazil, which was mostly unpleasant because the heat made it impossible to do anything outside.

Here are some options:

  • Heat banking: store heat during the summer in large reservoirs of water. Release it during the winter.
  • Heat trade: hot regions send heat to cold regions; both sides are happier.
  • Heat sequestration: there are huge pools of cold water about 1000m underneath the ocean surface, when your city is too hot, send some heat down there.
... (read more)
Reply
7James_Miller
Huge, thin mirrors in orbit over the equator that reflect light that would have hit the equator to the area of earth experiencing winter.
4roystgnr
There's a high-stakes variational calculus problem. For what seasonal temperature profile do we get the best long-term minimum for the sum of "deaths due to extreme cold" and "deaths due to tropical diseases whose vector insects are stymied by extreme cold".
2Lumifer
The problem is efficiency. Basically, it's (much) more efficient to control the temperature locally at small scale rather than transfer heat over large distances. There are some exceptions, e.g. geothermal can be very useful, see Iceland. I've seen mentions of trying to cool seaside cities with cold water pumped from the deep, but it's wildly more expensive than doing it the usual way.
1Marlon
Don't send heat underwater, it's a bad idea for everything that lives under there (and for us if we don't want these things going up). I'm curious though, how would you "send" heat ?
0Daniel_Burfoot
Not much lives 1000m under the surface. Also, the amount of heat that we would send is actually quite small compared to the heat capacity of the oceans. Water has 4000x higher heat capacity than air by volume. Transferring heat from a hot place to a cold place is really easy. In principle you can just connect them with a highly conductive material like copper. In practice even copper might not have enough heat conductance, so it might be better to pump either water or air from one place to another.
2root
Under the surface (for example, below the European continent) or in the deep seas? I'm not sure about the former but I'm quite confident that the following applies to the latter: My layman impression is that investigating lower altitudes becomes increasingly (perhaps exponentially) difficult the lower you go. Wikipedia also says that "Humans have explored less than 2% of the ocean floor" so I would disagree with your assessment of "not much lives 1000m under the surface". I'm honestly interested in how you came to that conclusion though - If you have an interesting and reputable text that refutes me, please share. I came to mine based on reading Wikipedia too much.
2Daniel_Burfoot
I should have said "not much lives below 1000m until you get to the ocean floor". Not much can happen in the deep ocean because light doesn't penetrate that deeply. The creatures that do live there have to rely on organic material falling slowly from the surface. From this article
0ChristianKl
https://en.wikipedia.org/wiki/Geothermal_heating is already a thing.
0SquirrelInHell
Don't you think it would, say, wreck most ecosystems on the planet?
2knb
It would only happen in areas of dense human habitation, which already wrecks the ecosystem. No net harm.
0[anonymous]
How about disrupting streams?
[-]James_Miller60

Meta discussion of the Crazy Ideas Thread

Reply
0Cariyaga
I don't think that's THAT crazy of an idea.
[-][anonymous]40

Let people with healthy feet sell their unwashed socks to people who have just finished fighting with pathogenic feet flora (if anyone has had a history of feet inflammation, they are prohibited to sell their socks.)

Reply
[-]gilch40

I've heard of cars powered by liquid nitrogen, since it boils at ambient temperature (even if the weather is below freezing), you can use it to expand a piston. The energy comes from the ambient environment.

Thermal equilibrium with outer space is about 4 Kelvin (due to background radiation). That's really cold. If we could make a large radiator exposed to open sky at night could we use it to produce liquid nitrogen? Not exactly, because the air itself can emit radiation. This is the greenhouse effect.

But would it be possible to coat the radiator with quant... (read more)

Reply
7Giovanni
In case you missed it and for future readers: It does indeed look like something like this is possible. At least two companies in the space: SkyCool and Radi-Cool. Here is an article covering mainly the former. I believes this qualifies the idea as No Longer Crazy™.
[-]lsparrish40

Orbiting landing tracks.

Payloads would be launched from earth with just enough fuel to loft them above the atmosphere and keep them hovering for a few minutes. Then they would electromagnetically couple to a long horizontal structure in low orbit, picking up velocity (or "losing" it, depending on the frame of reference) until they are orbiting at the same rate.

Electrically driven thrusters (e.g. vertical electrodynamic tethers which push against the earth's magnetic field) would be used to replenish the lost momentum. At any given time, the paylo... (read more)

Reply
[-][anonymous]30

Typical superheroes act like lymphocytes, while the bad aliens, against whom they fight, are antigens with no real planning beyond 'stake ground & multiply'; police in the stories are like interferon, and mass media are highly specialized, short lived antibodies, while the public shows general symptoms of inflammation if the situation spreads. Looping!days are malaria-like illnesses, apocalypses are...okay, there are really too many options...and mind control is, more or less, AIDS. Traveling between different universes is contagion.

Now, are there any stories sufficiently meta to be strictly epidemiological?

Reply
0polymathwannabe
Cabin in the Woods? Stranger than Fiction? Funny Games?
0[anonymous]
...I was probably wrong(?); could you explain what you mean, please?
3polymathwannabe
Those stories are not about something other than themselves and the rules/process/structure of storytelling. I felt they could match your request for something sufficiently meta.
[-]Gunnar_Zarncke30

A form of society that is based on a social contract where the rules (like the rules of e.g. democracy) are constructed in such a way that

  • individuals are (monetarily) rewarded if they take more responsibility (for other people basically) in a way that a doubled responsibility leads to a unit increase in reward (with a unit reward for taking responsibility for one self).
  • sub groups can make individual decisions and take local responsibility (and e.g. distribute the reward according to the rules)
  • there is an increase in power with group size but limited
... (read more)
Reply
4pianoforte611
I don't understand why you would want this. It doesn't take exactly X times as much effort to provide X times as much productivity, but its a way better approximation than a log scale. Is the goal to discourage commerce, and promote self sufficiency?
0Val
Another problem would be, that unless this system suddenly and magically got applied to the whole world, it would not be competitive. It can't grow from a small set of members because the limits it imposes would hinder those who would have contributed the most to the size and power of the economy. By shrinking your economy, you will become less competitive against those who don't adopt this new system.
2Dagon
If you have precise operational definitions of "responsibility", "reward" , "power", and "social product" that are amenable to public calculation, please share them. Getting agreement on the formulas should be a piece of cake if you can get agreement on the linear measures. If.
2Gunnar_Zarncke
The operational definitions that I used when I toyed around with this idea were: * reward is just monetarily income (or an equivalent thereof). I really should have made clear that this is imagined to be compatible with normal economics) * (amount of) responsibility is a) the number of persons you make decisions on behalf of with respect to (a subset of) the rules of the society. This responsibility can be agreed on (e.g. if you have a job where your decisions affect other people) or implicit (e.g. responsibility for yourself of your children) b) the monetary value of goods that you affect by your actions (again as agreed on or implicit), ** both can be fractional if responsibility is shared * social product is just gross national product ( https://en.wikipedia.org/wiki/Gross_national_product ) * yes, power is left vague. Choose any of (structural limit on) income or responsibility as above. I'm not sure which So8res was using in his quest but I assume that he also had operational definitions and also operational mechanics in mind.
2Dagon
Hmm. I don't think that works (or maybe you're just describing capitalism). If individuals have control over money, you can't have a societal design that contradicts their spending choices. * Currently, individuals are rewarded linearly for the amount*degree that they "take responsibility" (=sell products or provide service) for someone who has money to pay. This is a direct result of the operation of monetary transactions. * Currently, sub-groups can incorporate in order to make individual-like actions. * There is an increase in power with group size, but there's no simple definition of "power" that makes it feasible to calculate. I'm sure it is mathematically limited, though. * (optional) private property can be taken if powerful groups (governments usually) think they need it more than the owner.
2Gunnar_Zarncke
Capitalism allows unlimited private property with all the consequences for concentration of capital. If you don't want that you have to let go of something. Approaches that come to mind (and have been suggested) are * taxes on capital (which basically takes property away) * taxes on on income that approach 100% and thus limit income * limits for incorporation (anti-trust laws) * contracts that structure how income can be used (e.g. trusts are often highly structured in their spending) Basically I'm suggesting very general contracts that effectively change the linear reward effect you mention first to a consistent sub-linear one.
0Val
I fear some people will quickly learn how to game the system. No wonder our current society is so complicated, every time a group came up with a simple and brilliant way to create the perfect utopia, it always failed miserably. (also, try selling your idea to the average voter, I would love to see their faces when you mention "logarithm of total social product")
2Gunnar_Zarncke
Sure telling people that logarithms are involved will probably not help :-) Also oversimplification probably wouldn't work either. One key point is that - at a suitable level of abstraction - you can actually prove invariants of the system like limits to individual income/property/power. Invariants that you might or might not want to have.
0ChristianKl
You can prove limits to formal income/property/power but not to informal income/property/power.
2Gunnar_Zarncke
Yes. I kind of assumed that me leaving out formal definitions would prime people to think of plausible ones on their own - not that there are none or that I implied informal ones only. I gave some in the other comment. Or are you suggesting that a) there are no formal definitions or b) that there is no useful relationship between formal and informal definitions?
0ChristianKl
There are formal definitions but even through a country like North Korea doesn't like free markets it has Black markets in which people can earn income. In times of high taxes in France, rich people for example started to aquire a lot more art because art doesn't have an easily measureable and thus easily taxable value. Today it's possible to have assets in cryptocurrency.
2Gunnar_Zarncke
The idea is not to not have markets. The idea also isn't to take property away from people. They don't like it... The idea is to a) embed a better system in the existing society and b) to use basically existing and working means a systematic way, e.g. via contracts a la the viral GPL. If the system is better it will grow and eventually absorb the rest of society not by taking things away but by outcompeting. I'm asking to imagine such a system and what rules it works on. And don't tell me the current system is perfect.
[-]Good_Burning_Plastic20

(had the idea after seeing this)

Each person's vote should be weighed by their life expectancy given their age.

(ETA: I will downvote any comment in this subthread discussing the object-level issue of whether Britain had better stay in the EU, no matter how reasonable and insightful it is.)

Reply
4ChristianKl
That gives political power to those who calculate life expectancy. Should they get that power?
3gabrielrecc
This seems like a slippery slope. Minorities tend to have shorter life expectancies than whites, at least in the U.S. and U.K. Do their votes then count for less?
2Lumifer
At the bottom of that slippery slope is an ice floe.
0Good_Burning_Plastic
I'm not sure that's worse than what present-day Americans do.
0Lumifer
You do realize that getting onto that ice floe is not voluntary, right?
1username2
We can go even further: children's life expectancy could be added to their parents' voting weights. However most decisions are about the short term rather than the long term, and only very rarely life expectancy is relevant.
0username2
Obvious problems arise when aging is eliminated/fixed. But beyond that, why should one give those with least life experience & acquired wisdom the most vote? That seems entirely backwards. My initial expectation is that the reverse (one's personal voting weight vesting into increasingly higher influence over time) would lead to more harmonious societies, at odds with your suggestion.
[-][anonymous]10

Please take my consideration in moderation. I have a vested interest here. Also, I really love 80,000 hours and givewell, so I focus on the criticism for expediency: 80,000 highlights tobacco control in the developing world as one of the most important issues. However, I think they overestimate its neglectedness. As the tobacco atlas illustrates there is almost perfect symmetry between the research in the area and advocacy because you can see exactly what 'solutions' are required. Additionally, they underestimate the tractability of the problem. I don't ev... (read more)

Reply
0ChristianKl
Your post would be more convincing if you would provide an argument for why you think the cause is tractable. Insulting people isn't helpful. It's not a full investigation of the course area but as they say a limited look at the subject.
[-]bwasti00

I define intelligence as the ability to make optimal decisions to achieve some goal. The goal, clearly, is left undefined. This extends beyond the typical application of the word on humans, although I believe it fits nicely. A conventionally labeled intelligent person is capable of achieving conventionally defined "smart" goals such as performing well on tests and solving problems. However, things that are not seen as conventionally intelligent, such as the ability to distinguish between colors, would also fall under this definition.

One impli... (read more)

Reply
[-]gilch00

You've all seen the pendulum exhibit at the planetarium. Is it possible to use gyroscopes to extract the rotational motion of the Earth as a power source? Maybe you can use a vacuum and maglev bearings so you aren't expending energy to keep them spinning. You can use gears to trade torque for rotation speed. The available torque from the planet must be immense. Building such a device may be expensive, but then it's "unlimited" free energy with no carbon emissions, and, unlike most renewables, it has steady output.

Reply
0gwern
Googling, this turns out to have been discussed a lot more than I would have guessed. Apparently if it does work, even with very good ball bearing and gearing, you can't get more than a fraction of a watt, and that's worthless since such a scaled up gyroscope will break long before it pays back its cost, much less turns a profit.
0gilch
It's about what I figured. Energy is all around us, that doesn't mean its economical. I figure that a magnetic bearing will wear less than a ball bearing. How big does the gyroscope actually have to be for this to work? Can we just spin them faster? Why not an array of small ones? It might be cheaper to mass produce them. Also, the gear box was just proof of principle, you don't actually have to use gears. We could probably extract the energy more directly magnetically and trade volts with amps instead of torque with speed. And we can use superconductors to minimize losses there.
0ChristianKl
Superconductors need expensive cooling. In generally this seems both an expensive way to gather energy when we have relatively cheap solar panels and slowing down the earths rotation would likely be opposed by enviromentalists.
2gilch
Superconductors are themselves expensive, but are the cooling costs really that bad? I actually have another crazy idea for that. Slowing down the Earth's rotation is not a good argument against this idea. It would be a rounding error compared to the slowdown the Earth already experiences due to the Moon and tides. The day was 23 hours long at the time of the dinosaurs. Unfortunately, environmentalists might actually use the argument. They seem happy to oppose nuclear for stupid reasons. Cheap photovoltaics are coming, and they will probably use organic molecules rather than silicon. The problem remains grid scale storage. Photovoltaics only work when the light is on them. Solar can't be any cheaper than the cost of storage.
0ChristianKl
Cooling cost require liquid nitrogen. It's expensive. That's partly why MRI scans are expensive and why storing cyronics bodies is expensive.
0gilch
Liquid nitrogen costs something like $0.20 per liter, if you produce it at scale. If you buy if from someone else in small amounts it's naturally more expensive, but probably comparable to the cost of milk. My question isn't how much it costs to fill the tank in the first place, but rather how much boils off per unit time. A vacuum flask is a great insulator, so it might not be that much. If superconductors are necessary for enough efficiency to make this work, do we lose all our efficiency gains in cooling costs?
0lsparrish
It depends on the scale you are working at. A large body with no internal heat source can be kept cold over time at a lower cost because only the outside needs to be insulated. If cryonics were at the scale of a large cryogenic warehouse, it might be much less expensive.
0Lumifer
LOL
[-]ChristianKl00

I just started using the MessagEase keyboard on Android. It has a nice feature whereby pressing alⓒ gives you α and beⓒ gives you β.

On the other hand I have to press Alt +3B1 or Alt 224 to get α in windows and that's extremly difficult to remember. I think it would be great if somebody would write a program that also allows me easy access to such unicode characters on windows.

Reply
0gabrielrecc
I use Autohotkey on Windows for that purpose.
0ChristianKl
Do you have the script that you use on your Github?
0gabrielrecc
No; my script only contains the handful of unicode characters I commonly use, and is so idiosyncratic to me that it wouldn't be of much use to anyone else (mine includes autoreplacements for directories, email addresses I commonly type, etc.). But it's easy enough to make your own with whatever characters you use -- the syntax is simply ::text-to-replace::desired-replacement ::alpha::α ::em::— etc.
Moderation Log
93Comments