Are Financial Markets Too Short-Term?

Financial market prices embody info that helps others to make decisions. For example, firms decide activity levels based in part on their stock prices. Thus traders who add info to such markets do a public service, even if they do this for a private profit.

Such traders can choose to focus their info-collection efforts on “slow” info, which stays relevant for a long time, or on “fast” info, which is quickly forgotten. Many have said that such markets focus too much on fast info, relative to slow. In this post I will analyze this question. My tentative conclusion will be: yes, financial markets do indeed seem to focus too much on fast info. But first, let’s review the basics.

Each financial trade has an asset type, a buyer, a seller, a quantity, and a price. Each simple financial market trades one kind of asset, and its sequence of trade prices follows a random walk over time, a walk that reveals info about the value of that asset to observers. The expected price change variance during a time period is proportional to the amount of info revealed in that period.

Each trade happens via one trader first putting an offer into an “order book”, after which the another trader accepts that offer. While the act of posting a book order could reveal info to observers, it usually doesn’t. This is because a trader with substantial info prefers to instead profit from it by accepting a book order. If your info suggests that the price should rise, you buy, and if your info suggests that the price should fall, you sell.

However, the profits of traders who accept book orders come from the traders who posted those orders. So book order traders adjust their book prices to include the average info held by accepting traders. And competition typically moves book prices to where book traders make zero expected profits. There is a “bid-ask spread” between the “bid”, the highest book offer to buy, and the “ask”, the lowest book offer to sell. The size of this spread says how much info is expected to be embodied on average in each accepting trader.

However, some traders have little or no info. They instead want to trade for reasons other than profiting from info. If they could post competitive book orders, they should. But doing that well is hard. (For example, ~95% of book orders are cancelled before being filled.) So most low info traders instead accept book orders. Their trades lower the average info per trade, and thus allow traders with higher than average info to profit from their trades. These “fools” are the engine that drives the whole system.

For any given piece of info that a trader holds, they could profit more by trading a higher quantity at the same price. But those who make book orders foresee this strategy, and so their spread increases with order quantity; larger trades are presumed to carry more info.

As a result, a trader with an unusually big chunk of info prefers to reveal it more slowly over time, via a slower sequence of smaller trades (Vayanos, Kyle). And to avoid other traders noticing a pattern in their trades and jumping ahead to grab their profits, a trader who can find no other trades to hide among may need to make an apparent random walk of trades. For example, N2 trades on both the buy and sell side can hide N trades all on the same side.

So why not spread informed trades out over longer time periods? Because each piece of valuable info comes with a deadline. You can only profit from by telling a market about somethings that it will eventually learn in other ways. However, once many traders all know that many of them all have the same piece of info, then that info should be incorporated into the book order prices. Thus one can only profit by trading on such info before its everyone-knows-it deadline.

This duration-til-deadline varies greatly with info type. For example, slow info on future product fashions, or the success of innovation projects, may take years or decades to be revealed. In contrast, ~20% of trades are by “high frequency traders” (HFT), who typically trade on very fast info re prices in other markets. The deadline for the fastest HFT to arrive at a market with such other-market info is roughly when the second-faster HFT arrives. This is typically ~20-200 ns later for other markets at the same site, and ~50-500 μs for different sites (source: Kelvin Santos).

Thus five-year duration “slow” info is roughly a factor of a trillion to quadrillion times slower than HFT “fast” info. This huge dynamic range for info duration offers a big chance for duration effects to have big impacts. If there are problems with poor incentives re info duration, they could plausibly be really big problems. 

To evaluate whether financial markets focus too much on fast info, we should consider how social value, and also private trader costs and benefits, vary with info duration.

Let’s start with social value. As social value of info revealed to a market comes from its ability to influence decisions, decisions which are typically spread out across time, this value is roughly proportion to info size (i.e, price-change) times info duration. So, for example, if no relevant decisions are made using the market price in the few milliseconds duration of a high frequency trade, then the info in that trade induced zero social value.

Now let’s consider the private net revenue to be gained from a trade. As discussed above, that trade revenue is also proportional to info size times duration, at least for traders who have access to enough capital to support the required trading strategy, whose cost goes roughly as info size times duration.

How about trading costs? While there are fixed costs to design a trading strategy and arrange to implement it, and there can be mechanical marginal costs to execute a trade, the main other marginal cost is the opportunity cost of the assets used to make a trade. Any one asset can’t be simultaneously used to support an arbitrary number of arbitrary trades. The opportunity cost of these assets is also roughly proportional to info size times duration. (Yes, orgs that trade on margin and make many fast independent trades, may seem to face no opportunity costs of assets, but this is an illusion; they just have especially low opportunity costs per trade.)

So far all the factors we’ve considered have depended in the same way on duration; social value, trade revenue, and marginal trading cost all go as info size times duration. But a few considerations remain that depend differently.

For example, traders often do not have sufficient capital to fully profit from info that has a very large size times duration. In addition, long duration info apparently comes in larger chunks, which makes size and duration positively correlated. For example, an insight about whether some product innovation will succeed over the next decade is usually just a much bigger chunk of price-change-times-duration than is the last market price tick typically used by a HFT trader. This effect suggest insufficient attention to long duration info.

Finally, ambitious traders, and the systems that train and select them, prefer that traders show their abilities over many small fast trades, instead of over a few big slow trades. It is just not very useful to prove your trading abilities via finding and trading on info that takes decades to be proven right. This effect also suggests insufficient attention to long duration info.

Bottom line: while social value, trading revenue, and marginal trading cost all scale as price-change times info duration, the existence of large info chunks and the desire to prove trader abilities over career-sized durations suggests that financial markets pay too much attention to fast, relative to slow, info.

In my next related post, I’ll discuss how alternative trading institutions might mitigate this problem.

GD Star Rating
a WordPress rating system
Tagged as: ,

I See Stylists Everywhere

Cole: I see dead people.
Malcolm: In your dreams?
[Cole shakes his head no]
Malcolm: While you’re awake?
[Cole nods]
Malcolm: Dead people like, in graves? In coffins?
Cole: Walking around like regular people. They don’t see each other. They only see what they want to see. They don’t know they’re dead.
Malcolm: How often do you see them?
Cole: All the time. They’re everywhere.
The Sixth Sense.

In the movie The Sixth Sense, a boy Cole can see many ghosts, ghosts who are in denial about the fact that they are ghosts. I similarly see that most of our public intellectuals are in denial about the fact that they are mainly loved for their style, not their content. Let me explain.

A big reason that readers read (or listeners listen) is that reading transfers something from the mind of the writer to the mind of the reader, something that readers can then transfer to others if they themselves then write (or talk). I call that something the “content” of the writing.

All the other features of writing that can’t be as easily transferred, I call “style”. Readers may enjoy and admire the styles of things they read, but they can’t as easily add those styles to their own writing. Style is of course actually valuable, at least to some degree. Style can make content easier to enjoy, accept, and understand. The only question is how much effort to put into these different aspects of writing.

This difference between content and style is a continuum. For example, the main claim and argument of an essay is more content-like than are diversions and comments, which are less likely to be remembered or understood by readers. Subtext is even more style-like, as fewer readers can notice or repeat it. Subtle choice of vocabulary and sentence structure can be even more style-like.

(This way to define the distinction is unusual, but seems to me more concrete than most others you will find online.)

Content being more transferable than style has some big implications. Sometimes it is easy for writers who compete with you to read your writing and then put your content into their writing. In a way that few will complain about, because in these worlds there are not property rights in such content. In this case, you have less reason to invest in content. And this is in fact the situation for some kinds of intellectuals.

In contrast, this lack of property rights is less of a problem if you are a professional, paid by a client to write privately to them, on a topic of interest only to them, writings which they may then paraphrase to others.

In professions such as journalism, law, therapy, math, engineering, and business, people tend to adopt relatively standard and structured writing formats. For example, each essay is usually expected to clearly telegraph one main claim and argument. Many such professions also have their own specialized vocabularies and methods of inquiry, and their norms tend to prohibit many “fallacious” forms of argument. We tend to expect the best judges of professional writings are other professionals.

Such formats and habits tend to minimize opportunities for distinct personal writer styles and to maximize the ease with which readers can find and understand the distinctive relevant content of each case. That is, such professionals write as if they were interchangeable.

Such professions also often structure their relations to support property rights in writing content. Writers are either paid by specific clients to write on topics mainly of interest only to those clients, or professional norms of publication citation and priority help writers get credit for novel unique content.

Note that a memo to your boss is likely to be read only by that boss, who will paraphrase your content if it is passed up the management hierarchy. In which case your content will travel further than your style.

The writing style of many kinds of “public” intellectuals looks quite different from this. Such writers have much more distinctive personal styles, and allow themselves more digressions from a single main claim and argument. It is often unclear what exactly they are claiming, or which of their claims is central. And they often include examples, stories, and other discussions from which it seems hard to extract logical arguments. Such intellectuals also allow themselves a wider ranges of topics, methods of inquiry, and kinds of arguments. And for their writings we anchor more on the reactions of ordinary readers, rather than of similar professionals.

A straightforward interpretation of these differences is that these intellectuals emphasize style over content, relative to professional writing, because in their world it is much easier to own style, relative to content. Such intellectuals, however, usually resist such an interpretation of their own writing. If asked to explain why their writing looks so different, they point to the generality or value-orientation of their topics, or to their need to entice a wider range of more distractable readers.

Yes, they admit, they put a lot of effort into style. But this is only to make it easier for readers to enjoy, accept, and understand their content. They insist that the main reason that the world does, and should, attend to them, is their unique and valuable content. If one asks why then they don’t coordinate to create stronger property rights in content, they shrug, or suggest that the methods other professionals use won’t work for them.

I find the following observation especially telling. Many writers see themselves as specializing in content, and thus feel eager to team with writers who specialize more in style, to together make writings great in both content and style. But very few writers see themselves as style specialists, eager to team with content specialists. Almost all writers instead see themselves as having access to good enough content, thank you very much.

And thus we reach a Sixth-Sense-like situation. Just as in that movie, where a boy everywhere saw dead people, in denial about their being dead, I everywhere see stylists, in denial about their being stylists.

In that movie, sometimes one could convince a ghost that they were in fact a ghost. Could we similarly measure the degree of style emphasis of any given writer, and then use that to prove their style status to them?

One strategy would be to just pay other random writers to paraphrase essays, pay typical readers to read and rate those paraphrased versions, and then compare the popularity rank of original essays to that of paraphrased versions. (Using two different rank scales.) Writers whose actual essays rank higher than their paraphrased versions must be getting more of their popularity from style, which doesn’t get transferred in the paraphrasing process.

Of course it would be expensive to pay for these writings and readings. Maybe GPT will get good enough to do these things cheaper? Or maybe we could train such a system on a much smaller dataset of human reading and writing of paraphrased essays?

If the most popular (i.e., “top”) public intellectuals specialize the most in style, but still need good-enough content for their essays, that gives them a reason to look for content ideas by reading non-top intellectuals, who specialize more in content. Of course they might also need to read other top intellectuals, to track fashion trends in styles and topics.

This is my impression from attending social events with such top intellectuals. They don’t actually have more interesting ideas to discuss, compared to people a few ranks below them, and they spend most of their effort trying to track fashion trends and make connections. I noticed a similar thing when attending a group of investors interacting with a group of projects seeking funding; investors mostly tried to make connections and track trends in what the other investors were choosing, instead of evaluating projects themselves from first principles.

Some writers are especially interesting to top intellectuals, but are not top intellectuals themselves. These writers probably specialize more in good content. They will not get big book contracts, be invited as keynote speakers, or be assigned to top institutional positions. But they may still end up with outsized influence on what others come to think. Even if they don’t get much credit for that.

Note how content vs. style maps well onto experts vs. elites. Experts focus on content, while elites focus on style. Bosses are relatively elite, and they are often accused of presenting the content of their relative expert subordinates as if they had come up with it.

Added Jan5: Our norms generally tend to favor the production of things that would otherwise be underproduced. This predicts our norms favoring content over style in contexts with weak property rights on content. Which helps explain why we say that we favor it.

Many people think they find counter-examples to my definition of “content” vs. “style”, but they do not offer a concrete definition that accounts for all these examples. Seems to be just something folks “know it when they see it”, yet often don’t see the same thing.

GD Star Rating
a WordPress rating system
Tagged as:

Earth: A Status Report

In a universe that is (so far) almost entirely dead, we find ourselves to be on a rare planet full not only of life, but now also of human-level intelligent self-aware creatures. This makes our planet a roughly a once-per-million-galaxy rarity, and if we ever get grabby we can expect to meet other grabby aliens in roughly a billion years.

We see that our world, our minds, and our preferences have been shaped by at least four billions years of natural selection. And we see that evolution going especially fast lately, as we humans pioneer many powerful new innovations. Our latest big thing: larger scale organizations, which have induced our current brief dreamtime, wherein we are unusually rich.

For preferences, evolution has given us humans a mix of (a) some robust general preferences, like wanting to be respected and rich, (b) some less robust but deeply embedded preferences, like preferring certain human body shapes, and (c) some less robust but cultural plastic preferences, such as which particular things each culture finds more impressive.

My main reaction to all this is to feel grateful to be a living intelligent creature, who is compatible enough with his world to often get what he wants. Especially to be living in such a rich era. I accept that I and my descendants will long continue to compete (in part by cooperating of course), and that as the world changes evolution will continue to change my descendants, including as needed their values.

Many see this situation quite differently from me, however. For example, “anti-natalists” see life as a terrible crime, as the badness of our pains outweigh the goodness of our pleasures, resulting in net negative value lives. They thus want life on Earth to go extinct. Maybe, they say, it would be okay to only create really-rich better-emotionally-adjusted creatures. But not the humans we have now.

Many kinds of “conservatives” are proud to note that their ancestors changed in order to win prior evolutionary competitions. But they are generally opposed to future such changes. They want only limited changes to our tech, culture, lives, and values; bigger changes seem like abominations to them.

Many “socialists” are furious that some of us are richer and more influential than others. Furious enough to burn down everything if we don’t switch soon to more egalitarian systems of distribution and control. The fact that our existing social systems won difficult prior contests does not carry much weight with them. They insist on big radical changes now, and disavow any failures associated with prior attempts made under their banner. None of that was “real” socialism, you see.

Due to continued global competition, local adoption of anti-natalist, conservative, or socialist agendas seems insufficient to ensure these as global outcomes. Now most fans of these things don’t care much about long term outcomes. But some do. Some of those hope that global social pressures, via global social norms, may be sufficient. And others suggest using stronger global governance.

In fact, our scales of governance, and level of global governance, have been increasing over centuries. Furthermore, over the last half century we have created a world community of elites, wherein global social norms and pressures have strong power.

However, competition at the largest scales has so far been our only robust solution to system rot and suicide, problems that may well apply to systems of global governance or norms. Furthermore, centralized rulers may be reluctant to allow civilization to expand to distant places which they would find it harder to control.

This post resulted from Agnes Callard asking me to comment on Scott Alexander’s essay Meditations On Moloch, wherein he takes similarly stark positions on these grand issues. Alexander is irate that the world is not adopting various utopian solutions to common problems, such as ending corporate welfare, smaller militaries, and common hospital medical record systems. He seems to blame all of that, and pretty much anything else that has ever gone wrong, on something he personalizes into a monster “Moloch.” And while Alexander isn’t very clear on what exactly that is, my best read is that it is the general phenomenon of competition (at least the bad sort); that at least seems central to most of the examples he gives.

Furthermore, Alexander fears that, in the long run, competition will force our descendants to give up absolutely everything that they value, just to exist. Now he has no empirical or theoretical proof that this will happen; his post is instead mostly a long passionate primal scream expressing his terror at this possibility.

(Yes, he and I are aware that cooperation and competition systems are often nested within each other. The issue here is about the largest outer-most active system.)

Alexander’s solution is:

Elua. He is the god of flowers and free love and all soft and fragile things. Of art and science and philosophy and love. Of niceness, community, and civilization. He is a god of humans. … Only another god can kill Moloch. We have one on our side, but he needs our help. We should give it to him.

By which Alexander means: start with a tiny weak AI, induce it to “foom” (sudden growth from tiny to huge), resulting in a single “super-intelligent” AI who rules our galaxy with an iron fist, but wrapped the velvet glove of being “friendly” = “aligned”. By definition, such a creature makes the best possible utopia for us all. Sure, Alexander has no idea how to reliably induce a foom or to create an aligned-through-foom AI, but there are some people pondering theses questions (who are generally not very optimistic).

My response: yes of course if we could easily and reliably create a god to mange a utopia where nothing ever goes wrong, maybe we should do so. But I see enormous risks in trying to induce a single AI to grow crazy fast and then conquer everything, and also in trying to control that thing later via pre-foom design. I also fear many other risks of a single global system, including rot, suicide, and preventing expansion.

Yes, we might take this chance if we were quite sure that in the long term all other alternatives result in near zero value, while this remained the only scenario that could result in substantial value. But that just doesn’t seem remotely like our actual situation to me.

Because: competition just isn’t as bad as Alexander fears. And it certainly shouldn’t be blamed for everything that has ever gone wrong. More like: it should be credited for everything that has ever gone right among life and humans.

First, we don’t have good reasons to expect competition, compared to an AI god, to lead more reliably to the extinction either of life or of creatures who value their experiences. Yes, you can fear those outcomes, but I can as easily fear your AI god.

Second, competition has so far reigned over four billion years of Earth life, and at least a half billion years of Earth brains, and on average those seem to have been brain lives worth living. As have been the hundred billion human brain lives so far. So empirically, so far, given pretty long time periods, competition has just not remotely destroyed all value.

Now I suspect that Alexander might respond here thus:

The way that evolution has so far managed to let competing creatures typically achieve their values is by having those values change over time as their worlds change. But I want descendants to continue to achieve their values without having to change those values across generations.

However, relatively soon on evolutionary timescales, I’ve predicted that, given further competition, our descendants will come to just directly and abstractly value reproduction. And then after that, no descendant ever need to change their values. But I think even that situation isn’t good enough for Alexander; he wants our (his?) current human values to be the ones that continue and never change.

Now taken very concretely, this seems to require that our descendants never change their tastes in music, movies, or clothes. But I think Alexander has in mind only keeping values the same at some intermediate level of abstraction. Above the level of specific music styles, but below the level of just wanting to reproduce. However, not only has Alexander not been very clear regarding which exact value abstraction level he cares about, I’m not clear on why the rest of us should agree to with him about this level, or care as much as he does about it.

For example, what if most of our descendants get so used to communicating via text that they drop talking via sound, and thus also get less interesting in music? Oh they like artistic expressions using other mediums, such as text, but music becomes much more of a niche taste, mainly of interest to that fraction of our descendants who still attend a lot to sound.

This doesn’t seem like such a terrible future to me. Certainly not so terrible that we should risk everything to prevent it by trying to appoint an AI god. But if this scenario does actually seem that terrible to you, I guess maybe you should join Alexander’s camp. Unless all changes seem terrible to you, in which case you might join the conservative camp. Or maybe all life seems terrible to you, in which case you might join the anti-natalists.

Me, I accept the likelihood and good-enough-ness of modest “value drift” due to future competition. I’m not saying I have no preferences whatsoever about my descendants’ values. But relative to the plausible range I envision, I don’t feel greatly at risk. And definitely not so much at risk as to make desperate gambles that could go very wrong.

You might ask: if I don’t think making an AI god is the best way to get out of bad equilibria, what do I suggest instead? I’ll give the usual answer: innovation. For most problems, people have thought of plausible candidate solutions. What is usually needed is for people to test those solution in smaller scale trials. With smaller successes, it gets easier to entice people to coordinate to adopt them.

And how do you get people to try smaller versions? Dare them, inspire them, lead them, whatever works; this isn’t something I’m good at. In the long run, such trials tend to happen anyway, by accident, even when no one is inspired to do them on purpose. But the goal is to speed up that future, via smaller trials of promising innovation concepts.

Added 5Jan: While I was presuming that Alexander had intended substantial content to his claims about Moloch, many are saying no, he really just mean to say “bad equilibria are bad”. Which is just a mood well-expressed, but doesn’t remotely support the AI god strategy.

GD Star Rating
a WordPress rating system
Tagged as: , ,

Where You Stand Versus Sit

While this blog is called “Overcoming Bias”, I don’t recall explicitly addressing biases in a while. So let me revisit a bias which, though it is pretty clearly visible, most people don’t even bother to hide: where they stand depends on where they sit.

For example, their position on feminism and gender relations is predictable from their gender. Their position on redistribution is predictable from their generation. Their position on math vs words is predictable from their math vs word ability. Their position on a dispute between marketing and engineering depends on which division they sit in. And so on. If you give it some thought, you will notice that a lot of views are predictable, at least on average, from where people sit.

Yes, this is less of a problem for views on what is good for people like them, or what things look like to people like them. But most of us have a great many views on what is good in general, and what things are in general, views that are also predictable from who we are.

Yes, most of us can quickly point out exceptions, where our views go against that predicted by where we sit. But this is less about particular views and more about an overall pattern. (And it would be great if someone would set up a poll to show people just how well their views are predicted by where they sit.)

Yes, you choose some of where you sit. So you might claim that correlations between where you stand and sit are caused by your first choosing your view, and then choosing your place in the world. But this only works for a small number of views per place, as only a small number can plausibly have had strong influence on this choice. Thus such correlations become more or a problem the more different unrelated views are implied by your place.

We tend to be more tolerant of partisan views by experts and elites. That is, when experts disagree with non experts, or elites disagree with non-elites, we tend to take the side of the experts and elites. And thus are more okay with self-favoring views of experts and elites. But only because we are even less okay with those who disagree with experts and elite; be especially wary of your views being predictable by those features.

One reason we tend to tolerate this bias in ourselves is that we don’t mind showing allegiance to our associates, who tend to reward such loyalty. But ask yourself how biased you really want to be in favoring your associates.

This bias seems to me so pervasive that someone who had successfully rid themselves of it would likely also have rid themselves of a great many other biases. So this one seems well worth working on for that reason alone.

Added 1Jan: A Twitter poll asking people for their rank re this bias finds a pretty well-calibrated distribution. Thus people tend to accept and embrace this bias; they aren’t embarrassed enough by it to try to deny it.

GD Star Rating
a WordPress rating system
Tagged as: ,

More Data On The Sacred

To learn more about the sacred, I tried a few more Twitter polls. And one interesting meta datum I learned here is that few are curious about the sacred; when I asked for suggestions for more questions to ask, I got only one suggestion. Seems most are embarrassed by the sacred, and would rather pretend it doesn’t exist.

I asked about 16 sacred areas, somewhat different from my prior poll. These didn’t include politics, liberty, work, humor, or math. Re these 16 areas, I asked which you see as most sacred, which others in the world so see, what you’d like us to revere more, which has low conflicts with other sacred things, which have taken up most of your time, and which you see most emphasized in fiction.

Here are best fit priorities for these areas (max set to 100 for each area):

There are strong correlations between Self and Others (0.74), More (0.68), Time (0.34) and Fiction (0.36). Others only correlates with Time (0.38). All other correlations are <0.21.

These results suggest to me that there still remains much complexity in the sacred to understand. For example: As my followers see their priorities as different from others, what do those others say? Why are health/medicine, governance, war/nation, and law/police rated so low, when they seem to be treated so sacredly? Why aren’t other themes besides love given more emphasis in fiction? Why doesn’t a degree of conflict with other sacred things influence how sacred something seems? And as honesty/truth is top rated for wanting more of it, why don’t we have special rituals and holidays for it, as we do for other top sacred things? A similar question can be asked re innovation/inquiry.

GD Star Rating
a WordPress rating system
Tagged as:

On-Demand All-Topic Courts

There is not a natural alliance of contrarians. Instead, each contrarian group claims that it has been unfairly lumped in with the others. Sure, they say, most contrarians are wrong, but if you look carefully at our case, ours isn’t like those others; we happen to be right. So contrarian groups don’t like to associate with each other. Each group wants their associates to only associate with their contrarian claim, not with others.

But contrarians could usefully ally together to support institutions that better allow correct contrarians to prove their case to skeptical audiences. After all, while contrarians often elaborate their cases in great detail, typically few outsiders have much incentive to consider those details, leaving audiences to rely mainly on the raw fact that most others dismiss their case while ignoring their details. A process which doesn’t do much to distinguish correct from incorrect contrarians.

As many have noted, a general solution to this problem is on-demand all-topic courts. If you pay enough to such a court, and give it a claim, it will have high quality people give substantial but fair attention to the details of that claim. And if such a court also had a sufficient reputation for being neutral and canonical, then audiences might well listen to it in the exceptional cases where its verdicts supported particular contrarians. And the prospect of such vindication might cause contrarians to pay them.

For each case, such a court could collect three teams: pro and con advocates, and a jury. The pro argues for the claim, the con argues against it, and the jury listens and issues a verdict. If so, the main design questions re such a court are how to ensure sufficient neutrality and quality in these teams. As there is no such court yet, a new court in this space has a reasonable hope to become the canonical court in this space, if it can be seen as sufficiently neutral and qualified.

If we can trust the jury to be fair, qualified, and attentive, then the pro and con advocate teams don’t need to be fair; they just need an incentive and tendency to win. So I propose this process: (A) publicly offer most of the new-case fee as prize to the winning advocate, (B) advocate candidates declare their willingness to argue particular sides, (C) choose pro and con teams via decision markets estimating each candidate’s win chances, and (D) choose the jury via a status app.

The simple status app I outlined in my last post only offers general eliteness. So a status app variation capable of saying who is higher status re particular topics would be even more useful. We might also pay jurors in part in bets that a future randomly-created jury with a larger prize will agree with their verdict. (Perhaps we could even pick jurors using decision markets estimating such a future verdict agreement.)

We’d like to avoid imbalanced advocate incentives. If we create prediction markets on the various possible jury verdicts, then the court could bet in those markets to acquire assets ready to pay off advocates, and ensure that both sides have the same expected payoff, given the market prices just before the court process begins. With this, the two sides would have similarly strong incentives.

And that’s my proposal for an on-demand all-topics court, which could give correct contrarians a better shot at distinguishing themselves from other contrarians. Other than how to substitute for prediction markets if those are illegal, the main remaining ambiguity I see here is how to design and field a relevant status app; I discussed that in my last post.

Note that contrarians with only a small budget to fund a prize might want to use an audit lottery to create a small chance of a much larger prize. Then could might create prediction markets estimating jury verdicts conditional on such a large prize case happening. Such market prices might well suggest that their case is not as weak as many presume it to be.

GD Star Rating
a WordPress rating system
Tagged as: ,

Status App Concept

In my last post I suggested that we prefer institutions of this form:

Masses recognize elites, who oversee experts, who pick details.

However, our existing methods for doing that first step, masses recognizing elites, seem rather limited. One simple method is to inherit a stable social consensus on the relevant weights to give different status markers, such as wealth, birth family, test scores, endorsement of prior elites, or winning between-elite fights. If we agree on such weights, we might quickly agree on who has how much status. Especially when we pick just one of these markers as our main marker of status.

A second method is to use our ancient more complex, opaque, and instinctual human methods of gossip, displays, fights, and other social tricks to come to a shared consensus on who is higher status. As most human communities do in fact come to rough consensus relatively quickly on relative status judgments, humans clearly do have such mechanisms, even we don’t understand them very well. But such gossip, displays, and fights are often very expensive.

A third method is elections, wherein masses choose between elite political candidates, mostly based on the advice of other elites. While electoral systems usually only only have the capacity to set the status of a tiny number of top officials, those few top officials can sometimes set the status of more others below them.

However, it isn’t clear that any of these methods actually give the masses that much influence over elite policy, or even over the relative status of elites. Nor do they seem that great at preventing coalitions of elites from installing themselves as unaccountable dynasties. Nor are they obviously great at picking the best people to be elites. Can we do better?

In this post I will outline a concept for a more fined-grained and decentralized approach: a status app. Though I haven’t figured out all the details, I’m posting this partial concept to entice you all to help me think about the remaining design issues. And then maybe implement something.

But first, let’s get clear on the relevant standards for evaluating such a proposal. Our other systems for agreeing on status induce great costs, and also suffer from strategic gaming, and a great many personal biases and agendas. Thus a new system doesn’t need to eliminate all such problems to be an improvement. It might be good enough if it just does better re some problems, and not worse re other problems. Furthermore, the first version of such a system needn’t be better than the status quo, if we can use trial and error to improve it, to eventually make a much better system.

Okay, here is my proposed concept. In a new status app, the core action is this: random triples of people X,A,B are selected, and then person X is asked which of the pair A, B they more respect, at least re elite social roles. Their answer is a “status bit”.

In my simplest reference design, the status app just fits a simple statistical model to all of the bits it has seen, a model with parameters that include each person’s current status, and the current info (vs noise) level of each person re their status bits. If these parameter estimates are made made public, the world could use them as input to many other social processes.

For example, the app might compute a status Elo score for each person based on their “wins” vs. “losses” in each of their status bit “contests”. Each person’s info score could then be a time-weighted measure of how well their their status bits predict changes to target Elo scores in the time period after their bits.

Now let’s consider some design issues that might drive us to modify this reference design.

The first issue is what triples X,A,B to use to create status bits. Yes one could choose them completely randomly. But to get bits that better help the app to estimate parameters, it would make sense to slant the triples toward X whose bits are more predictive, and toward A,B pairs whose status estimates are closer to each other. Also, toward triples X,A,B who have closer relations and more similarities to each other. And especially toward situations where X actually sees A and B interact, or sees them act in closely related contexts. Especially situations where status judgments are usually and naturally made.

When we can categorize the context type C for each status bit, it would make sense to have an info parameter for each such type, so that the expected error (squared) for each bit depends on both the particular person X as well as the context type C.

However, the more control that X,A,B have over which triples are evaluated when, the more they will try to game these choices, and the more inclined X might be to sacrifice their info score in order to reward or punish associates. There is thus an open question re what kinds of status judging contexts C to include in this system, and who to let cause or veto each one. Is it sufficient to just include an error adjustment parameter for each different context type C?

A second issue is that even a decent statistical model of these status bits will likely have known errors, inducing participants to try to game them. This certainly happens re Chess Elo scores. If we believe that eventually sufficient data will be collected to make such errors small, so that earlier large errors are mostly temporary, then it may be sufficient to create prediction markets on future Elo estimates, and use current market prices as our best status and info estimates, instead of stat model estimates.

But if we can’t trust model errors to fade away with time, then we might instead want prediction markets that pay out based on random future status bits, using context types that are especially hard to game (as in this post). This approach forces market traders to suffer higher risks, but is safer re model estimate errors.

A third issue is who is allowed to see what status bits. One extreme is where everyone can see them all, while the opposite extreme is where only bit creators can see them. The closer that X,A,B, are to each other, the more risk there is of inducing problematic behaviors by X,A,B toward each other when they can the details of such bits. But the more people who can see more all the bit details, the better they might be able to correct model estimate errors in prediction market prices.

A fourth issue is whether we can create a decentralized implementation of such an app concept, so that we don’t have to trust some center who might lie about or distort such a system.

A fifth issue is that while asking people who they generally respect more seems a very direct way to elicit general status judgments, what we might really want as data are actions where people reveal who they actually fear (for dominance) or seek to emulate (for prestige). But could we really find a set of actions where (a) we could reliably extract relative status judgments from those acts, without too many other confounding effects, (b) the set of actions covers a wide enough range of status aspects to allow the estimation of general status, and not just one narrow aspect of status, and (c) such actions can be observed often enough to give sufficiently accurate estimates of individual status? This seems hard.

A sixth issue is whether and how to give higher status people more weight in judging relative status. Will their info estimates naturally be higher in the basic stat model, or do we want to favor them more than would be done by this analysis, and if so how?

A seventh issue is that sometimes data may be of the form of (X,Y,…) ranking many people (A,B,…,N). Can this be reduced to many bits of the form of ranking (A,B), or does a stat model need to handle this differently?

An eighth issue is how to merge different kinds of status bits. That is, if X picks the higher of A and B several times re several different aspects of status, should the different kinds of status bits be analyzed separately, or would it be useful to estimate their correlation and use those to estimate each kind of status for each person?

I’ll add more issues here as I think of them.

GD Star Rating
a WordPress rating system
Tagged as: ,

Elites Must Rule

While I’ve spent much of my life doing institution/mechanism design, I’ve only lately come to see that, at least on prestigious topics, most people want relevant institutions to take the following ideal form:

Masses recognize elites, who oversee experts, who pick details.

While experts are known by other experts to be knowledgeable and skilled regarding particular topic areas, elites must be widely seen by the masses as having connections and features that are admirable and appropriate for leadership and governance roles. (More on experts vs elites.)

For example, with democratic governance the voter masses elect politician elites, who appoint mixed-elite-expert agency heads, who oversee expert agency personnel, who choose details. And debates on who should get to vote can be seen as debates on voter eliteness.

Long ago, it was seen as sufficient for the masses to recognize the natural eliteness of kings and aristocrats, who descended from or choose each other. You might think we are past that now, but academia and medical regulation both work this way. That is, we are all supposed to recognize that top academics and doctors are sufficiently elite to be trusted to run those orgs, and to pick their own successors, all without substantial outside oversight. And actually, this is how regulation works in many prestigious areas in many nations; the elites of each area are given a free hand to run those areas, and choose their successors.

If one sees the typical rich investor as sufficiently elite, then one can accept investors overseeing CEOs who oversee middle managers who oversee line workers, in a progression from relatively elite to relatively expert. The main reason that people object to this arrangement is that they are reluctant to see investors as sufficiently elite; they prefer instead to have government officials decide which firms get funding, and to have workers elect their managers.

Most people don’t like direct democracy, and dislike it more more when ordinary voters have more influence over the proposals on which to vote. And most people aren’t actually that comfortable with legal juries, unless jury choices are greatly limited by elite judges, and advised by expert lawyers. The U.S. uses juries today much less than it used to, and most nations have little use for them. It seems that most people instead want something closer to the ideal form described above.

Back in 2003 my Policy Analysis Market project hit the news, and was immediately killed. While the loudest complaints against it, of sabotage and price manipulation, had little basis in fact, this oft mentioned criticism was solid: such prediction markets would somewhat displace prestigious intelligence elites with more ordinary people. Markets producing better decisions via more accurate estimates was not seen as sufficient justification.

In the CIA today they tolerate internal prediction markets, but only under the rule that market estimates are never to be cited in official reports, which are the coin of prestige in that realm. Having prestigious reports cite prediction markets would let low level CIA experts somewhat displace CIA elites.

Wariness of many of my other institution ideas, such as tax career agents, life maintenance orgs, and crime vouchers, might also be attributed to their more overtly distrusting elites in various ways, and displacing the usual elites somewhat by for-profit firms and financial investors. And these are parties that many see as insufficiently elite. Producing more cost-effective outcomes is not seen as sufficient justification.

While I will continue to try to persuade people to weaken this constraint, and to consider institutions that rely less on our just trusting prestigious elites, I will now also try to take this seriously as a design constraint, and design institutions that conform better to the ideal form described above. I’ll start in my next post.

GD Star Rating
a WordPress rating system
Tagged as: , ,

Less Talk Context

Here’s a long-term trend I don’t recall hearing much about: over time, talk has been losing its non-talk context. That is, over time listeners have known less about the context of speaker talk.

Though animals have quite limited languages, they often manage to say what they need to say. And shared context helps with that. For example, each animal may see a lot about where they each are and what they are each doing.

Human language allows for a lot more to be said via our words, and even more via style, such as tone, pacing, etc. But in private talk, we humans still rely heavily on context to make ourselves understood. This context includes not just where we are and what we are doing, but also our histories, relations to each other, and what we have heard about each other via gossip.

Humans also often distinguish between “text” and “subtext”. The literal meanings of our words often differ from, and even contradict, what we say via style and context. As quotable texts are usually designed internally to make subtext deniable, seeing style and context help a lot for inferring subtext.

How much context humans have to interpret words has always varied. For example, when humans addressed larger groups, less could be inferred from their relation to particular audience members. And low context situations have increased over time. For example, with rising population densities, individuals more often talk to relative strangers.

While the introduction of writing has allowed the exchange of letters between friends, writing also allowed a single speaker to address many diverse people across space and time. Furthermore, schools and mass media have greatly encouraged many to spend a lot of time reading such low-context writings. In the last few decades, social media has gone further, encouraged many ordinary people to spend a lot of time writing in a lower-context mode as well. And very recently, large language models trained on big datasets of such public talk have been able to mimic this low-context talk style impressively well.

We humans change our talk in many ways to deal with lower context. For example, speakers add more expressive language, and distinctive talk styles, in order to create stronger packages of listener expectations. And because listeners consistently seek subtext, they more aggressively infer “implicatures” from what speakers say. For example, we feel more free to attribute motives according to speaker demographics, or by “political” associations.

Low context writers seeking to avoid accusations of subtext often use a defensive low-emotional “bureaucratic” or “official” “classic” style. This style admits of no motives other than telling you simply and directly what the writer sees. And as this is style of much of the text on which large language models have been trained, and as the sponsors of such models seek to avoid criticism of their models, these models also tend to admit to no other motives. Also, as classic official talk tends to be “socially desirable”, avoiding cynical appearances, these language models also tend to be reluctant to suspect low motives for human behaviors.

GD Star Rating
a WordPress rating system
Tagged as: ,

We Sweat Big Stuff Badly

Don’t sweat the small stuff; and its all small stuff.

Compared to unimportant decisions, for moderately important decisions we tend to do more thorough practical decision analyses. This is mainly because we try harder. Yet when we get to our most important decisions, our decision analyses tend to be less explicit, thorough, or practical. This is in part because we tend to see such issues as more symbolic, and more sacred.

For example, we think more carefully about if and how to repair a car, compared to a broken pencil. Yet when it comes to thinking about repairing our bodies, our thinking tends to get less thorough and analytic. We would rather just trust our prestigious doctor, and those who conferred prestige on them, and not think about the subject. It seems that we mainly use medicine as a way to show that we care, instead of as a way to get well. And as a result, people who we randomly induce to get more medicine are not on average substantially healthier. That’s a pretty big fail.

We handle small conflicts all the time, and we put more effort into handling moderately important conflicts. Yet our biggest conflicts get handed over to a quite dysfunctional legal system. (A system few are willing to think of redesigning.) We each assume it works well until we personally have to deal with it. Then we try to just trust our prestigious lawyer. We show little interest in lawyer track records or incentive contracts; we just don’t want to think about the topic.

We are eager to connect with each other socially, and in a low key informal contexts where the same people regularly come into contact with each other, people will in fact chat, socialize, and get to know each other. In in those processes, they will successfully tend to get more contact with people they like better. But even the loneliest people are quite reluctant to directly approach strangers. We’d rather stay lonely than to risk a clear public rejection.

At home, we are generally capable of doing modest chores, and of learning the skills needed to do them. And we put more effort into learning the chores that matter more. But at school and on jobs, we often self-sabotage and put in low effort on the biggest choices, in order to make failure less of a signal of what we could accomplish if only we’d try harder. I’ve done this many times in my life. Often I was most productive when I’d play hokey from one “should do” project to work on another “not supposed to be doing” project.

Our disfunction re big work choices seems especially dramatic in the case of people who choose hobbies and video games with tasks, and related performance feedback, are quite similar to those on real jobs. They say they’d rather not work for money because they hate the prospect of being criticized or looked down on for poor job performance.

Yes many, perhaps even most, people slowly “mature” over time into becoming better able to “own” and practically consider big decisions. But even so, most of us still aren’t very good at this. So as a lot rides on such big decisions, how can we get people to better sweat the big stuff?

One approach is to pair people with incentivized expert advisors who can push particular decisions, more or less gently. These advisors can be more calculating and dispassionate, as such decisions don’t loom as large for them personally. And less individual more communal societies often do pair people with associates who advise them, though it is often unclear how expert or well-incentivized is such advice. Fans of regulation also suggest that regulators would do well in this advisor role, though critics of regulation have doubts.

This issue is one of the reasons I’m so interested in designing and testing better personal agents. Like tax career agents, life maintenance organizations, or legal liability vouchers. Such agents can arguably be both expert and well-incentivized to advise well. Making us more comfortable with encouraging ordinary folks to rely more on their advice, instead of on their own broken decision-making.

Added 18Dec: Thought some claim that standard decision theory is inadequate for big decisions, I’m not at all persuaded of that.

GD Star Rating
a WordPress rating system
Tagged as: