Original Lesswrong thread here.
Original tweet here:
Unlinked market with shorter timeframes here: /Joshua/when-will-we-know-that-any-past-ufo
1,000
3.00Welcome to the market where we think we’re trading on the most fundamental questions of life in the universe, but really we’re trading on whether we get to borrow fake money or not
@Irigi This really makes me doubt my usual automatic approach to Manifold market prices as an epistemological signal, a probability estimation from a well-calibrated tool. What a change in probability estimation we're experiencing, and how little it has to do with any actual epistemological update! What other market dynamics am I missing that influence market probabilities and make them less well calibrated?
@wadimiusz Correct. Market prices are not probabilities - it is more accurate to say that they can be used to estimate probabilities, but they are not unbiased estimators. This is for many reasons, one of which (discount rates) is very important here and I discuss in this prior comment
@jack Doesn't the platform's calibration (and each individual user's calibration) treat prices as probabilities? I.e. if stuff I where buy YES at 30% resolves YES in 20% of the time, then I'm overconfident, but if it resolves YES 40% of the time then i'm underconfident etc.
@wadimiusz That much is true, but that is only a first approximation and missing a ton of detail. Some notable examples:
Discount rates: if the market price is 2% and I believe it's 1%, and it resolves in 2 years, then that's a very low ROI and it's not worthwhile for me to bet NO On the other hand, if I believe it's 3%, then that's a huge ROI and it is worthwhile for me to bet YES. When you look at this across the market, it tends to skew the price to be closer to 50% than the average prediction of the market participants.
Risk: Market prices are adjusted for risk, they are not pure probabilities. See Risk-neutral probabilities for a description of this concept.
@jack Wait, but the standard pitch for prediction markets is something along the lines of "either the market is calibrated better than you are and already takes all information you have into account, or you can get rich (or in our case, rich in fake internet money) with no downside". That's how @ScottAlexander presented it, anyway! Does this mean I don't get to use this pitch at parties anymore? (I'm being cheeky, but it's true that I sometimes advertise prediction markets at parties, and that's how I usually go about it...)
@wadimiusz That is all basically true, but it doesn't mean the market will be exactly correct. If the market is mispriced too much, it will be profitable to correct it. But how much is too much varies a lot, and can easily be 5% or 10% or more.
Also, nothing I'm saying is specific to prediction markets. It applies to stock/bond/option/etc markets too. Prediction markets are really just another financial instrument, similar to option markets. They also are predictions on the values of various financial things, and those predictions are usually pretty accurate. If you interpret the price as literally a prediction, you won't be far off. But that isn't actually accurate and you could easily be 10% off, say.
@wadimiusz People here are biased to say this is just as accurate as real money markets but they are wrong. They are wrong no matter the topic of the market, including ones they consider themselves experts in (like AI), or ones they consider themselves the only non-biased people (like politics.)
The real money markets are better, every time. And the price here would be far closer to 100% if this were a real money market.
@DavidBolin no, there are real money UFO markets on polymarket and they show much higher discount rate effects than the mana market, because of the expectation that mana loans will come back - they all have much shorter timeframes but still price UFOs at several percent
@DavidBolin @jack is 100% correct here… the timeframe is just too long. Why would you invest your money in a bond returning <5%/year?
@SteveSokolowski loans are probably being reintroduced in the next couple of months, so the expected rate of return for YES has gotten better
@benshindel If that's the case, I'm probably going to leave Manifold, or at least curtail my usage.
I spent a lot of mana here and at other markets and made a lot more, and then they diluted the value of the mana. Now, they're changing the rules again.
This is why I never build commercial products that rely upon a third-party provider. They can decide to deny you service at any time and all the money you sunk into using their system is lost. Here, I'm fortunate that I never participated in any sweepstakes or lost any real money, but surely they must recognize the damage this will do to their reputation among existing customers.
I certainly will never bet any real money on this platform if they go through with a major change like this.
@IsaacCarruthers It's more of a breaking point than anything else. This particular change will cost will wipe out about 100,000M or so of mine, but the previous changes devalued it already. It's just one thing after another until I start to get tired of it.
@SteveSokolowski This market is absurd and would benefit from a spending cap per person. The fair market price should be at 99%, but I will probably never get any benefit from betting it up.
@SteveSokolowski I'm a little baffled how you see this as wiping out Ṁ100k. You, too, will receive loans on your investment in this market which you can invest elsewhere profitably. If the probability goes up and decreases the current value of your investment, you can just hold and we'll all see who's right in the end.
I guess if loans are based on current investment value then the price moving up means you'll receive smaller loans for a lower rate of return. On the other hand, NO shares will be cheaper if you want to keep buying.
@SteveSokolowski if it “wipes out 100k of value” that means that you made 100k of bad bets! If you’re correct, then this market will eventually resolve NO and you’ll make tons of profit! If you’re incorrect, then you will rightly lose all of that mana! That’s how prediction markets work.
@SteveSokolowski also it’s basically never optimal to bet 90% of your net worth in a single market (least of all one priced at about 10% or whatever)
Under no circumstances is that an optimal market strategy.
@benshindel Steve here was linked a video easily debunking some basic facts about the alien book he was proclaiming as unimpeachable evidence a month ago and pretended he couldn’t watch it. I think it’s most likely he’s optimizing for attention and engagement, not for mana. It gets way more interaction and attention to throw everything into something silly like this and use it to lend legitimacy to picking fights with everyone, including the platform itself, than spread it around to multiple reasonable markets with varying confidence levels. We see that in multiple threads here.
If I’m right more generally, one may expect some performative shenanigans in the future with his market about getting called out as a troll and/or his market about using chatGPT and Claude as his lawyers against Wells Fargo.
@benshindel Yes it is. It's an optimal strategy when you have extreme conviction that an asset is dramatically mispriced.
That's how I made so much money putting 90% of my net worth into bitcoins in 2013; it was a dramatically mispriced asset. In the case of this market, there are two related reasons this market is dramatically mispriced:
The evidence for the existence of non-human intelligence is overwhelming.
Most of the bettors in this market have not actually read any books on the topic they are betting on, and they admitted as such in the comments. One even made a market as to whether he should perform some research into the topic and decided not to.
If I were betting against people who were actually informed on the topic and the market still wasn't moving, that would inform my position. Remember, there are studies that show when scientists are given just 20 hours to perform research into this topic, the percentage of them who believe it should be investigated jumps from approximately 30% to 70%.
I bought DG stock and it crashed; I'm not mad at that, as I performed research and thought that dollar stores would excel in a recession. The NYSE didn't randomly change the price of the stock, like Manifold did with the price of mana. You can't just change the rules in the middle of the game.
@jcb Perhaps you misunderstood, and for that I apologize. The loans didn't wipe out the 100K M; the loans were responsible for a much smaller loss.
The loans would reset the market to the previous state if the company hadn't caused mana to be so ridiculously devalued for no reason at all. The mana was lost when they devalued it.
Anyway, these changes are too discouraging and I cancelled all my orders - I don't care anymore.
@SteveSokolowski a Kelly criterion calculator implies that if you’re making optimal bet-sizing, you believe this market has about a 75% chance of resolving in your favor. I guess that’s your prerogative, but that seems irrationally high. Even if there WERE a world-shattering origin for UFO’s (which there isn’t), the odds of it coming to light by 2028 seem much lower than 75%.
(And fwiw, this market has never gone anywhere NEAR 25%)
@benshindel I agree with you that the revelations are unlikely to come from the government. I expect them to come from o1-lol or its successors within 6 months, which will clearly be "AGI" by any reasonable definition and will be able to reason through the huge dataset on these issues.
I actually don't expect the market to get to 25%. The evidence is already pretty overwhelming - certainly beyond the threshold warranting significant resources spent investigating - and as I said, people both here and in the general public simply aren't willing to read up on the topic. I expect that science will continue to waste money on SETI instead of performing studies on this for a long time to come.
The market will be fairly valued when science starts researching the topic and the average person is informed about how much information is already known about UFOs, whether anything new is known at that time or not.
But the market doesn't have to go to zero for me to make a lot of money, just as I never believed that bitcoins would be worth $100,000 like some people were preaching they would be someday. All that needs to happen is for something to occur that causes people to actually read about the topic - like the publicity from the November 13 hearing or the upcoming documentaries on the various streaming platforms.
@SteveSokolowski I don’t understand what you mean by “the market doesn’t have to go to zero for me to make a lot of money”
This seems to imply you’re counting on the greater fool theory to kick in so you can cash out?
@benshindel No. I'm counting on people becoming less of fools. The studies show that when people read about the topic, they are much more likely to believe it worthy of investigation. I'm betting on awareness of the topic spreading, encouraging people to read about it, and those people betting the probabilities closer to what they should be.
@SteveSokolowski I'd of course be willing to believe that when people read conspiracy theories they were previously not exposed to, they are more likely to believe them. If you give 1,000 people a book with flat earther propaganda, they'd be more willing to believe that too. I don't think that should provide evidence that flat earther conspiracies are meaningful or true. But that's also not how prediction markets work! With 2 month loans back on Manifold, I'd probably be willing to buy about 100k worth of shares at ~90% in this market, giving me an expected R.O.R. of about 5%/month.
@SteveSokolowski you're betting this because of a short term keynesian beauty contest that more conspiracy theorists will join manifold? More evidence that this market should just resolve YES right now so the conspiracy theorists lose mana.
@benshindel It’s also worth noting that someone believing something is worth investigating is not remotely the same as believing it is aliens. This is an obvious misrepresentation typical of his trolling.
@benshindel Most conspiracy theories have some parts of them that are true. In this topic, about 90% of it is trash and 10% of it has extremely strong evidence.
I'm not going to debate people who haven't actually read about the topic. If you want to debate specific government reports, sightings, technologies, religious events, etc, feel free to contact me. Otherwise, my stance on talking about with people on this topic remains the same as it was a few months ago. Become informed about what you're debating if you want to debate it.
If that's the case, I'm probably going to leave Manifold, or at least curtail my usage.
Win win!
@SteveSokolowski I agree that sooner or later AGI will have the capability to reason its way to the truth on this topic. Unfortunately all the big models are also brainwashed into the approved groupthink on topics like this, so they’d probably just spit out the normal “there’s no evidence” and “it’s important to note that science..” etc etc.
@SteveSokolowski why do you think I’m not informed on this topic? Why do you assume I haven’t read about it? If your bar on “being informed” is spending a large fraction of my life reading conspiratorial websites, then no, I do not meet that bar. But I’ve read enough about it to come to an informed opinion.
Become informed about what you're debating if you want to debate it.
At a certain point, debate is tiresome. The whole point of prediction markets is we can let the bets do the talking.
If you think this question resolves NO with probability >20% (implied by you buying NO @ 83%), the re-introduction of loans should benefit you. You get to buy shares at a cheaper price and with leverage.
The only way that loans hurt your position is if you don't actually think this has any chance to resolve NO, in which case you were mainly betting on greater-fool theory (knowing that YES buyers can't correct the price due to time/risk discounting) and should understand the risks of such a position.
@benshindel I guess my disappointment with your comments is your use of labels like "conspiracy theory." I saw this in the Presidential election market where you get a lot of supporters from both sides who use simplified terms rather than discussing the issues.
You probably indeed do know a lot about the topic; but I hope you can see that your intelligence doesn't come through when you use those terms. If I stood on the other side of this issue, my post would be something like "the smoking gun showing It's unlikely that any UFO is attributable to non-human intelligence because the UAPDA's terminology avoids using the term 'alien,' and Chuck Schumer's speech could be seen as a bipartisan approach to uncover crime because Mike Rounds didn't say X in his speech." Or, another argument might be "some of the abduction narratives recorded in the book might be influenced by the psychologist's suggestions, given that the study wasn't double-blinded."
It's great that you know a lot, but I would love to actually talk about the facts. Those two particular arguments might be compelling, but I could point out the flaws if you talked about the facts. Using blanket terms such as "conspiracy theory" shows closed-mindedness and (I believe) ultimately exposes one to the possibility of ending up holding erroneous beliefs, as I have started to find recently when I started looking into Biden's health and what Trump's policies actually are.
@SteveSokolowski believe me, I have a pretty high bar when it comes to calling something a “conspiracy theory”. For instance, I would not call belief that Oumuamua is alien in origin a “conspiracy theory”. But if you want me to get specific, as an example:
-All claims by Grusch that the US gov is harboring knowledge of alien artifacts or biologics are completely false and belief in that is indeed a CONSPIRACY theory because for it to be true, not only would it have to defy logic and evidence on multiple levels, but also it would require the government (and many other parties) to be engaging in a CONSPIRACY to cover it up. In this case, it’s quite literally a “conspiracy theory”.
Even if you believe in these conspiracy theories, you shouldn’t really have any grounds to object to them being called such as it would quite evidently require a massive conspiracy to cover up and suppress these findings!
@benshindel But what specific evidence are you pointing to that would be defied?
Again, this is the main issue I have with your and others' arguments - specificity. You haven't pointed to the specific evidence that leads you to your conclusions.
By the way, I do happen to think that reading about things people call conspiracy theories is worthwhile. I read a lot about QAnon and its beliefs to get an understanding of what it is. When I read about it, my evaluation was that most of it was false, but I came away with an understanding of why people could get caught up in it. That helps me understand why some of the people who participated in the coup did the things they did, and helps me when I interact with die-hard Trump supporters who support him solely because of his false statements.
I read about UFOs for the same reason - to try to understand the theory. Through that research, and because I kept an open mind, I changed my opinion and decided that it's more likely than not that the evidence supports many of the claims. I can point to specific evidence, like certain sightings that have the five observables, along with the statements of specific high-ranking government officials and the actions of lobbyists, and Jacques Valleé's "control system" hypothesis, along with what he found in regards to the motivation behind a few obviously false hoaxes in France.
If you want to point to specific evidence that defies logic, then that's fine; I'll be glad to talk about it. I just don't understand why this specific topic causes so many people here to so strongly express the other point of view without actually reading in depth and making their own opinions about it.
@SteveSokolowski lol this thread was fun to read but regardless of why you think this could resolve NO, for sure you do understand 95% of the people won't read this massive amount of text and would just think "this question is stupid" and buy YES, right? I would honestly buy it up but I don't trade markets with long resolution times