R&Ds human systems http://aboutmako.makopool.com
mako yass
Karma: 3,967
The usual thought, I guess. We could build forums that’re sufficiently flexible that they could have features like this added to them without any involvement from hosts (in this case I’d implement it as a Proposal/mass commitment to read ‘post of the day’s, and the introduction of a ‘post of the day’ tag. I don’t think this even requires radical extensibility, just the tasteweb model), and we should build those instead of building more single purpose systems that are even less flexible than the few-purpose systems we already had.
Do we know whether wolves really treat scent marks as boundary markers.
Some confusing things about wolf territoriality is they frequently honestly signal their locations through howling while trying (and imo failing?) to obfuscate their number in the way that they howl.
Not a coincidence, there are practical reasons borders end up on thresholds. A sort of quantization that happens in the relative strength calculation. Two models:
Simple model: You could say the true border can be defined in terms of the amount of time or effort it takes to get there from the cat’s houses. It takes a certain amount of time and effort to get over a fence, so if the true border is (from tuxedo’s house) between distance tud + yd and tud + yd + fd, then the border in practice will end up being exactly on the fence, because you can’t put a border halfway up the fence, or the situation would look the same if you did.
A more accurate model: The border is measured in terms of how hard it is to defend a space from being used by the other side. I’d guess that cats become vulnerable to attack when they mount a fence (same as humans crossing a river), either coming or going, so extending your territory beyond the fence is difficult. If your strength is higher than the amount of strength it takes to defend everything before the threshold, but lower than the amount it takes to cross the threshold and then defend some on the other side, then the border will be exactly on the threshold.
I’m pretty sure I’d predict no for 1. Cats don’t seem to care about that stuff.
For 2, I’m not sure, if there were a hole in the fence, I’d expect confrontations to happen there because that’s a chokepoint where a cat could get through safely if the other one wasn’t standing on the other side, and maybe the chokepoint is a vulnerability threshold, too. Chokepoints are thresholds for projectile combat (because when you come through the defender sees you immediately but you don’t spot them until they start shooting), cats may be partly characterizable as stealth projectiles.
Also worth noting is that, eg, dogs, they engage in “boundary aggression” at things like fences, but experiments show that they’re doing it for the love of the game. If you remove the fence, hostilities cease. Cats may have some of this going on as well. They may on some level enjoy yelling and acting tough while being in no risk of having to actually fight.
3: Yeah, but because it makes the relative strength calculation harder. A fence is a blessed device that allows cats to get a good look at each other without engaging. I wish humans had something like that. (A hole in a fence may also be a good device for this)
A schelling point is an arbitrary default choice converged upon without communication when agreement is needed more than correctness. A territorial border between animals is an extremely non-arbitrary result of often very thorough tests of relative strength and communications of will. Animal borders are opposite to schelling points.
Borders between human territories are pretty arbitrary, we don’t really have the kind of bounded conflict that can produce a relative strength estimate any more (some of us used to), and most of us engage in antinductive commitment races by propagandising mythic histories about the legitimacy of our land claims (I don’t believe in that shit though, personally). The present order truly seems to be satisfied with schelling points for borders, it doesn’t matter what you choose, as long as we can agree, and never disagree, and whatever we agree about is the true border.
But animal borders aren’t arbitrary, they’re constantly renegotiated. The negotiations may be partly tacit, but there’s nothing whimsical or symbolic about the outcomes. The animals know where the resources are, they know how much they want them, they know their neighbors, they know how often the neighbors come to check their border so they can estimate the amount of pressure they face, and they know the risks of getting into a fight with their neighbor, so they’re able to make really pretty rational calculations to decide where the borders are.
I don’t know what you’re asking. The answer is either trivial or mu depending on what you mean by specific form. I think if you could articulate what you’re asking you wouldn’t have to ask it.
The VNM axiom isn’t about road trips, a utility function is allowed to value different things at different times because the time component distinguishes those things. You aren’t addressing VNM utility here. You’re writing about a misunderstanding of it that you had.
You die if you have VNM cycles. A superior trader eats you (People feel like they can simply stop communicating with the sharps and retire to a simple life in the hills, but this is a very costly solution and I’d prefer to find a real one). You stop existing. This is kind of a much more essential category of instrumental vice than like “I don’t equate money to utility” type stuff (which I wouldn’t call a vice).One criticism of decision theory that you could explore is that many practical philosophy enjoyers would find it difficult write utility functions that compose scripted components (like “I want A, then B, then C, then A”) with nonscripted components (“I will always instantly trade X for Y, and Y for Z”), that we may need higher level abstractions on top of the basics to help people to stop conflating ABC with XYZ… but… is it really going to be complicated? That one doesn’t seem like it’s going to be complicated to me.
What does seem difficult is expressing constrained indifference about utility function changes. Something that seems to be common in humans (eg, I’m indifferent to the change/annihilation of my values if it’s being done by beautiful and cool things like love, literary fiction, or reason, but I hate it if it’s being done by ugly or stupid or hostile things.) and is needed for ASI alignment (corrigibility), but it seems tricky to define a utility function that permits it. (though again I don’t know whether it turns out to be tricky in practice)
A utility function that enjoys moving between those places isn’t the same as a utility function with cycles, which would trade unlimited time money for tickets to them that it never cashes.
The argument against this is that is also going to be somewhat instrumental in flavour but more along the lines of like, that’s a known attracter that few who matter want to be in.
In a world that has ASI, a much better way of maintaining the integrity of the audit system by building it to be intelligent enough to tell whether it’s being fooled, and with a desire of its own to stay neutral. Which I guess is like being multistakeholder, since you both will have signed off on its design.
But in such a world, the audit system would be a feature of the brain of the local authorities. You would co-design yourselves in such a way that you have the ability to make binding promises (or if you’re precious about your design, co-design your factories in such a way that they have the ability to verify that your design can make binding promises (or co-design your factory factories to …)). This makes you a better/viable at all trading partner. You have the option of not using it except when it benefits you. But having it means that they can simply ask you whether your galaxy contains any optimal 17 square packings, and you send them an attestation that no when you need to pack 17 squares you’re using the socially acceptable symmetrical, suboptimal packings, and if it has a certain signature then they know you weren’t capable of faking this message.
You really don’t want to lack this ability.
But starting from the outset with little shared resources and (obviously) allowing establishment of shared projects using their resources by agreement between the stakeholders doesn’t seem much different from some de novo process of establishing such shared projects with no direct involvement of individuals
You’re speaking as if we’re starting with strict borders and considering renegotiating, for most of the resources in the universe and also on the planet this is not the case, ownership of space is a taboo, ownership over ocean resources is shared, at least on the nation level. It’s as if humans have shame, sense the absurdity of it all, and on some level fear enclosed futures. I think shared ownership (which is not really ownership) is a more likely default, shared at least between more than one person, if not a population.
But to the point, I don’t think we know that the two starting points lead to equivalent outcomes. My thesis is generally that it’s very likely that transparency (then coordination) physically wins out under basically any natural starting conditions, but even if the possibility that some coordination problems are permanent is very small, I’d prefer if we avoided the risk. But I also notice that there may be some governance outcomes that make shared start much less feasible than walled start.
Ownership is the ability to fully exclude others from, or if you wish, dispose of, an object. Ownership is an extremely dumb negotiation outcome for any object larger than a sparrow. It’s something that humans think is fine and eternal because of how dumb humans are. We simply aren’t able to do better, but better deals are easily imaginable.
As an example of why you wouldn’t want to pay the premium (which would be high) of full ownership over a galaxy: If you have sole ownership of something, then you can exclude others from knowing what you’re doing with it, so you could be running torture simulations in there, which would bother other people a lot, just because it isn’t in my yard doesn’t mean it’s not affecting my utility function, so you would have to pay an insane premium for that kind of deal. You’d prefer to at least cede a limited degree of ownership by maintaining constrained auditing systems that prove to your counterparties that you’re not using the galaxy to produce (much) suffering without proving anything else, and they’d be willing to let you have it for much less, in that case.
And in a sense we’re already part of the way to this. You can buy an animal, but in a way you don’t completely own it, you aren’t allowed to torture it (though for the aforementioned humans being dumb issues you can still totally do it because we don’t have the attentional or bureaucratic bandwidth to enforce those laws in most situations in which they’d be necessary). If you mistreat it, it can be taken away from you. You could say that this weaker form of ownership is simply what you meant to begin with, but I’m saying that there are sharing schemes that’re smarter than this in the same way that this is smarter than pure ownership. Lets say your dog looks a lot like a famous dog from an anime you’ve never seen and never want to see. But a lot of other people saw it. So they want to have it cosplay as that for halloween, while you don’t really want to do it at all. Obviously going along with it is a better negotiation outcome, society in theory (and sometimes in practice) would have subsidised your dog if they had an assurance that you’d fulfil this wish. But it wont, or can’t afford to. So you don’t do it. And everyone is worse off, because of how extraordinarily high the transaction costs are for things as stupid as humans.
By episode 8, I find it to be more of a depiction of extinctionist boddhisatvism than a depiction of a credible mode of AI, though the former is still sometimes relevant around these parts.
The virus is a form of violence that converts humans into very nice and happy people at the modest expense of estranging them from most of their desires, and in so doing eventually destroying most of what they used to value. A rejection of any passion strong enough to move a person to defend the things they love from those who’d tread on them, to an extent that they can no longer really claim to have love.
but thought that people won’t have a good reason to fill out their trust weights
Yeah, I notice that using a transitive quality as the endorsement criterion, and making votes public, produces an incentive for a person to give useful endorsements: Failing to issue informative endorsements would indicate them as not having this transitive quality and so not being worthy of endorsement themselves.
We can also make it prominent in a person’s profile if, for instance, they’ve strongly endorsed themselves, or if they’ve only endorsed a few people without also doing any abstention endorsements (which redistribute trust back to the current distribution). Some will have an excuse for doing this, most will be able to do better.I wonder if it’s good to pre-fill the trust weights (e.g. based on AF upvotes history), to make it easier for users (and motivate those who strongly disagree with their defaults)
True. Doing that by default, and also doing some of the aforementioned abstention endorsements by default, would address accidental overconfident votes pretty well.
(Also, howdy, I should probably help with this, I was R&Ding web of trust systems for a while before realising there didn’t seem to be healthy enough hosts for them (they can misbehave if placed in the wrong situations), so I switched to working on extensible social software/forums, to build better hosts. It wasn’t clear to me that the alignment community needed this kind of thing, but I guess it probably does at this point.)
Collective LessWrong Value:
If everyone who used LessWrong would pay the same amount you do for the website, how much would you pay? (In USD)Should probably say “Per year”
Also it’s a very tricky question because it seems to assume that we can start charging people without decreasing the number of users, in which case the price should probably be extremely high, higher than any online service has ever costed, due to the fact that it’s almost never possible to charge what a public information good, or its impacts, are worth (it’s worth a lot).
How month long vacations would you trade for a new sportscar? If you’d trade months of vacation for one sportscar, write 2, if you’d trade one month of vacation for two cars, write 0.5.
Many typos here. Also I hate it. Which sportscar. Why not just give a dollar value. My mind compulsively goes to the tesla roadster which’ll probably have cold gas thrusters and so is likely to value a lot more than the average sportscar. The answer will also be conflated with how much people like their work. Some people like their work enough that they’ll have to give a negative answer, or they might just answer incorrectly based on varying interpretations of what a vacation is, can you work during a vacation if you want to? I’d say not really, but I’m guessing that’s not what you intended.
(previously posted as a root comment)
Where do you live? It’s conceivable that a suit actually does mean these things where you live, but doesn’t in the bay area. Some scenes/areas just don’t expect people to dress in normative ways, they’ll celebrate anything as long as it’s done well.
It’s important to separate the plan from the public advocacy of the plan. A person might internally be fully aware of the tradeoffs of a plan, while being unable to publicly acknowledge them, because coming out and publicly saying “<powerful group> wouldn’t do as well under our plan as they would under other plans, but we think it’s worth the cost to them for the greater good” will generally lead to righteous failure, do you want to fail righteously? To lose the political game but to be content knowing that you were right and they were wrong and you lost for ostensibly virtuous reasons?
I think Reddit tried something like that; you could award people “Reddit gold”, not sure how it worked.
It didn’t do anything systemically, just made the comment look different.
You need to have a way to evaluate the outcome
What I plan on doing is evaluating comments partly based on expected eventual findings of deeper discussion of those comments. You can’t resolve a prediction market about whether free will is real, you can make a prediction market about what kind of consensus or common ground might be reached if you had Keith Frankish and Michael Edward Johnson undertake 8 hours of podcasting, because that’s a test that can/may be run.
Or you can make it about resolutions of investigations undertaken by clusters of the scholarly endorsement network.
The details matter, because they determine how people will try to game this.
The best way to game that is to submit your own articles to the system then allocate all of your gratitude to them, so that you get back the entirety of your subscription fee. But it’d be a small amount of money (well, ideally it wouldn’t be, access to good literature is tremendously undervalued, but at first it would be) and you’d have to be especially malignant to do it after spending a substantial amount of time reading and being transformed by other peoples’ work.
But I guess the manifestation of this that’s hardest to police is; will a user endorse a work even if they know the money will go entirely to a producer who they dislike especially given that the producer has since fired all of the creatives who made the work.
I’d expect the answer to not be apparent to an outsider, reading the literature, but I’d expect people who are good at designing those sorts of systems to be able to give you the answer quite easily if you ask.
I wonder if this is a case of gdm optimising for the destination rather than the journey. Or more concretely, optimising for entirely AI-produced code over coding assistants.
I think it’s not a fluke at all. Decision theory gave us a formal-seeming way of thinking about the behaviour of artificial agents long in advance of having anything like them, you have to believe you can do math about AI in order to think that it’s possible to arrest the problem before it arrives, and also drawing this analogy between AI and idealised decision theory agents smuggles in a sorcerers apprentice frame (where the automaton arrives already strong, and follows instructions in an explosively energetic and literal way) that makes AI seem inherently dangerous.
So to be the most strident and compelling advocate of AI safety you had to be into decision theory. Eliezer exists in every timeline.