I took the survey and mostly enjoyed it. There are some questions that I skipped because my answer would be too specific and I wanted to keep the ability to speak about them without breaking anonymity.
I also skipped some questions because I wasn’t sure how to interpret certain words.
I don’t have much to add. But I think this is a very well done post, that it has a nice size in scope, and that it is about a good and useful concept.
Why not make it so there is a box “ask me before allowing other participants to publish” that is unchecked by default?
Thank you for your work. I am often amazed by the effort you pour in your regular posting, and I see you as among the highest value contributors to LW.
Two minor nitpicks:
As opposed to, as Solana states here, saying ‘e/acc is not a cult [....] technologically progressive meme.’
Maybe format this part like other quotations? I was confused for a second there.
Memes do not declare anyone who disagrees with them on anything as enemies, who they continuously attack ad hominem, describe as being in cults and demand everyone treat as criminals. Memes do not demand people declare which side people are on. Memes do not put badges of loyalty into their bios. Memes do not write multiple manifestos. Memes do not have, as Roon observes, persecution complexes.
Meh. Groups of people sharing a common meme certainly do all these things. I see little point in arguing the precise semantics of “movement” and I do not particularly think Solana’s message is honest. But I would like to register that I don’t see any intrisic contradiction in “X is a rare meme that turns those who integrate it to their thinking into rude unthinking fanatics, but it did not create a movement yet”.
I think this somehow misses the crux for the problem of induction. By redefining “justification” (and your definition is reasonnable), you get to speak of partially justifying a general statement that has not yet been fully observed. Sure.
But this doesn’t solve the question. The question, in your example, would be closer to “On what basis can I, having seen a large number of swans being white, consider it now more likely that the next swan I see will be white”?
I think it already exists to some extent, at least in France. We have “leboncoin” which I think is similar to Craigslist. Many offers are very cheap and the software is decent, though not great. The giver has to deal with the hassle of taking pictures, making a public offer, and then coordinating with the taker; so in exchange the taker gives them a token amount of money. Seems fair. I truly think that many offers on leboncoin are put there because people want others to benefit from what they no longer need. I also think I saw some offers in the past that were fully free (listed for 1 euro, with a comment saying they are free).
You say
Did craigslist pave the way despite becoming a cesspool of overly expensive crap and deceptively listed ‘$1’ ads for businesses, that require you to drive 8 miles to pick it up while someone side-eyes you suspiciously from their driveway and waits for the venmo to go through?
But my experience with leboncoin in France has always been about nice people and polite conversations. Maybe there is a cultural difference?
I sometimes think that it would be great to have a more comprehensive system to facilitate giving and reusing all kinds of things across society. It would notably be great to have a system that handles storage until a taker is found, pickup/delivery, quality check, and cleaning in exchange for a small fee. We also have that in France for furniture, it is called Emaus. You give your couch to the charity, they clean it and ensure it is in decent shape (I think they do bedbug screening) then they put it up for sale and deliver it, all for a very reasonable price. In addition it even provides jobs for people in need who might often be “noncompetitive” and unlikely to find jobs otherwise. Emaus is a charity, so they don’t care: the jobs are part of the goal. Their website is quite bad, there is no accessible database of items on offer and they don’t even do a good job at clarifying their services and policies. So you need to go there and see for yourself. Why don’t they have a good website/app? I don’t know but I can guess. Not a priority, their volunteers are often old and don’t know how to code, it would take more work for each donation if they had to populate a database, everything gets sold eventually so they think better software would change nothing.
In short: Craigslist/leboncoin is already decent software for donations where you pay the donor a few bucks for taking the time to organize the donation, and people use that software to that end. Maybe there would be benefits to an app that only organizes donations and nothing else, but the gain seems minimal. It would make sense to have a third party simplify the logistic for a fee, and that has been done for expensive items (couchs). For it to work with smaller stuff you would probably need some economy of scale, so it might be harder to make it work. Maybe there is an opportunity there. You certainly would need good software: I might go to Emaus to see if they have a nice couch, but I won’t go for a blanket if I cannot check in advance that they have a nice one.
This looks interesting. I will come back to this post later and read it if the math displays properly.
I have become less sceptic about the ability of western government to act and solve issues in a reasonable timeframe. In general, I tend to think political actions are doomed and are mostly only able to let the statu quo evolve by itself. But recent relatively fast reactions to the evolution of mainstream AI tools have led me to think that I am too cynical on this. I do not know what to think instead, but I am now less confident in my old opinion.
Here is a description of that metaphor, for those who don’t know.
I have heard of filters and ultrafilters, but I have never heard of anyone calling any sort of filter a hyperfilter.
Oops, my bad. I re-read the post as I was typing to make sure I hadn’t missed any explanation. That can sometimes cause me to type what I read instead of what I intended. I probably interverted the prefixes because they feel similar.
Thank you for the math. I am not sure everything is right with your notations in the second half, it seems to me there must be a typo either for the intersection case or the superset one. But the ideas are clear enough to let me complete the proof.
Thank you for this. it looks like a good first contact with hyperreals.
Two nitpicks:
Ω=(1,2,3,ldots). --> I think you forgot a “\” here and it is messing your formatting up.
It is not clear in the post why we use a hyperfilter, rather than just the set of all infinite sets.
I am also curious about this association. Is it just that in your experience these traits are correlated?
I agree that deep canvassing would be interesting. I am also curious about the famous experiments in which forcing people to smile (by holding a pen in their mouth) makes them more likely to appreciate something. Though I don’t know if there are already many replication studies for those.
But many as an economist you would consider this to be too out of your field?
No, I mean the price at which that party is indifferent between making the deal and not making the deal.
I think that’s the same thing? By “highest price” I meant “the highest price the buyer is willing to pay”. That’s the turning point after which the buyer dislikes the deal and before which the buyer likes the deal.
Yeah, and I’m trying to make that difficult for humans to do.
I understand but I fail to see that this attempt works. It seems to me that in many / most real cases (for which I have a reasonable estimate on the other’s best price) it is in my interst to lie if I know that the other is filling the form honestly. If that is correct, then the “honest meta” is unstable.
Just in case: I assume that by “best price” you mean “highest price” rather than “estimated fair price”.
If so, I only need to have some information on it to be incentivized to lie. In the example above I only use the information that the buyer is willing to pay two units above the fair price. The kind of example I use doesn’t work if I have no information at all about the other’s best price but that is rare. Realistically, I always have “some” estimation of what the other is willing to pay.
If we take a general Bayesian framework, I have a distribution on the buyer’s best and fair price. It seems to me that most/all nontrivial distributions will incentivize me to lie.
I am not so sure the incentives are properly aligned here. Let’s assume I am the seller. In the extreme case in which I know the highest price accepted by the buyer, I am obviously incentivized to take it as my own limit price.
And I think this generalizes. If:
there is a universally agreed true fair price FAIR_PRICE, like an official market value
the buyer is still filling the form honestly
I know the buyer to have a highest price above FAIR_PRICE + 2 then I can easily get FAIR_PRICE+2
Of course this requires some information on the other negotiator, but I do not see this as unreasonable.
Could you clarify in which situation this is meant to incentivize people to fill the form honestly?
I guess. But I don’t know of any real-world transactions where it’s expected that people keep their word on something like this
I think there are two points worth raising here:
If someone accepts to precommit to the result of this negotiation and then, when the website outputs a price, refuses to honor it then I probably do not want to trade with them anymore. At the least, I would count it as though they agreed to a price and refused to honor it the next day.
You only need to keep a solid precomitment yourself to avoid falling prey to the strategy above.
The obvious issue in my eyes is that few people would agree to use this kind of tool anyway, especially for a nontrivial transaction. But in a society (or simply a subculture) that normalizes them it would probably no longer be true that people don’t consider their precommitment to the tool as binding.
The obvious exploit is to lie and then negotiate “normally” if the tool fails to make a deal in your favor.
The website says:
In order for both participants to have the correct incentives, you must both commit to abide by that result and not try to negotiate further afterwards.
So this strategy fails against people that keep their word.
Maybe you would accept this paper, which was discussed quite a bit at the time: Emergent Tool Use From Multi-Agent Autocurricula
The AI learns to use a physics engine glitch in order to win a game. I am thinking of the behavior at 2:36 in this video. The code is available on github here. I didn’t try to run it myself, so I do not know how easy to run or complete it is.
As to whether the article matches your other criteria:
The goal of the article was to get the AI to find new behaviors, so it might not count as purely natural. But it seems the physics glitch was not planned. So it did come as a surprise.
Maybe glitching the physics to win at hide and seek is not a sufficiently general behavior to count as a case of instrumental convergence.
I won’t blame you if you think this doesn’t count.
When I downvote a comment it is basically never because I want the author to delete that comment. I rarely downvote comments already bellow 0, but even when I do it is not because I wish the comment was deleted. Instead, it mostly means that I dislike the way in which that comment was written and thought out; that I don’t want people to have that style / approach when commenting. This correlates with me disagreeing with the position, but not strongly so; and I try to keep my opinions about the object topic to the agree/disagree voting.
I don’t know how representative I am of the Lesswrong population in that regard, but I at least think most people who downvote a comment would prefer for it to stay undeleted; if only to make past discussions legilible.