This post raises some reasonable-sounding and important-if-true hypotheses. There seems to be a vast open space of possible predictions, relevant observations, and alternative explanations. A lot of it has good treatment, but not on LW, as far as I know.
I would recommend this post as an introduction to some ideas and a starting point, but not as a good argument or a basis for any firm conclusions. I hope to see more content about this on LW in the future.
Firstly, on a historical basis, many of the greatest scientists were clearly aiming for explanation not prediction.
In all of your examples, the new theory allowed making predictions, either more correct than previous ones (relativity, astronomy) or in situations that were previously completely un-predictable (evolution). Scientists expected good predictions to follow from good explanations, and they were in large part motivated by it.
Wiener, on the other hand, is saying it doesn’t matter what explanation you choose if all explanations yield the same prediction, in a particular field of study or experiment. And you don’t need explanations at all if they can’t ever yield different predictions (in any possible experiment). That’s a different statement.
I think that taking prediction to be the point of doing science is misguided in a few ways.
This seems to be just a matter of definitions. Scientists are human beings, they have a wide variety of interests and goals. You can label a more narrow subset of them “science”, and then say that some of what they’re doing “isn’t science”. Or you can label everything they tend to do as “science”, because it tends to come together. But the question “what is the real point of doing science?” is just a matter of definition.
I’ll open the discussion for the broadest definition of ‘value drift’ - ‘changes in values over time’.
‘Good’ and ‘bad’ are only defined relative to some set of values.
A simplistic (but technically correct answer): if you had values A, and then changed to have different values B, from the viewpoint of A this is bad *by definition*, no matter what A and B actually are. And from the viewpoint of B it’s good by definition. Values are always optimal according to themselves. (If they’re in conflict with one another, there should be a set of optimal balances defined by some meta-values you also need to hold.)
A more complex and human-like scenario: you’re not perfectly rational. Knowing this, and wanting to achieve a certain goal, it might be useful to “choose” a set of values other than the trivial set “this goal is good”, to influence your own future behavior. Just as it can be instrumentally rational to choose some false beliefs (or to omit some true ones), so it can be instrumentally rational to choose a set of values in order to achieve something those values don’t actually claim to promote.
A contrived example: you value donating to a certain charity. If you join a local church and become influential, you could convince others to donate to it. You don’t actually value the church. If you were perfectly rational, you could perfectly pretend to value it and act to optimize your real values (the charity). But humans tend to be bad at publicly espousing values (or beliefs) without coming to really believe them to some degree. So you’ll get value drift towards really caring about the church. But the charity will get more donations than if you hadn’t joined. So from the point of view of your original values (charity above all), the expected value drift to (charity + church) is an instrumentally good choice.
Loyalty, authority, and fairness are also about other people. A lone person can’t be loyal, authoritative, or fair; you have to be those things to someone else.
And, as I’ve been saying, Harm/Care is also about the conduct of the individual: do you harm others or care for them?
“I can’t learn the material for you” as opposed to “if you want to climb Mt Everest, you have to do it for yourself rather than for someone else”.
I’m not sure I understand the difference, can you make it more explicit?
“I can’t learn the material for you”: if I learn it, it won’t achieve the goal of you having learned it, i.e. you knowing the material.
“I can’t climb the mountain for you”: if I climb it, the prestige and fun will be mine; I can’t give you the experience of climbing the mountain unless you climb it yourself.
The two cases seem the same...
if people care about others being pure, it seems they can just as easily care about others being caring. And that we should think about people trying to observe the norm of caring and making sure others do, rather than trying to care effectively. Is that right?
Yes, that’s what I think is happening: people observing norms and judging others on observing them, rather than on achieving goals efficiently or achieving more. Consequentially, we want to save everyone. Morally, we don’t judge people harshly for not saving everyone as long as they’re doing their best—and we don’t expect them to make an extraordinary effort.
And so, I don’t see a significant difference between Harm/Care and the other foundations.
I did not mean to misrepresent what lawyers do (or are allowed to do). I noted they are restricted by lawyer ethics, but that was in a different comment than the one you replied to. Yes, absolutely, they not supposed to lie or even deliberately mislead, and a lawyer’s reputation would suffer horribly if they were caught in a lie.
I’m not sure I understand people who aren’t OK with ethical lawyers, as a concept. Is there something they would like instead of lawyers? (See: my other comment.) Or do they feel that lawyers are immoral by association with injustice—the intuition of “moral contagion” (I forget the correct term) that someone who only partially fixes a moral wrong, is worse than someone who doesn’t try to fix it at all?
Harm/Care is unusual among the foundations in that it’s other-directed. The goal is to help other people, and it does not especially matter how that occurs. [...] In contrast, the other foundations centre on the moral actor themselves. I cannot be just, loyal, a good follower, or pure for you.
It seems to me that Harm/Care isn’t as different as you say. Native (evolved) morality is mostly deontological. The object of moral feelings is the act of helping, not the result of other people being better off. “The goal is to help other people” sounds like a consequentialist reformulation. Helping a second party to help a third party may not be efficient, but morality isn’t concerned with efficiency.
In contrast, the other foundations centre on the moral actor themselves. I cannot be just, loyal, a good follower, or pure for you.
I could say: yes, I can be just *to* you, loyal *to* you, a good follower *of* you. And pure *for* you too—think about purity pledges, aka “save it *for* your future spouse”.
In all these cases, morality is about performance—deontology—rather than about accomplishing a goal. But each case does have an apparent goal, so our System 2 can apply consequentialist logic to it. Why do you treat Harm/Care differently?
I think my definition of rhetoric is the same as OP’s: namely, the art of shaping words or a speech to be beautiful, moving, convincing, or otherwise effective. How to best verbally convince others of an idea: I think that’s a useful term.
In particular the OP referred to dispositio (concise, addressing the right points) and pronuntiatio (body language and delivery).
I’m not convinced this is true.
I’m not sure what exactly you’re not convince of. That speech is much more effective when its form is liked as well as its object level claims?
I don’t think that’s true. Lots of people are bothered by this. Maybe you’re right, maybe a majority is unbothered, but this is interesting only to the extent that it doesn’t embody a larger pattern of what proportion of people care about injustice.
I agree that most people are bothered by anything they perceive as injustice. But if they don’t know a way to make things better, or what things being better would look like, then they tend not to blame e.g. lawyers for participating in the system and being good at it.
Is there a better way of doing things, that lots of people would prefer to be the case? Not just “I wish judges applied the law fairly and for Justice”—then you might as well wish for people not to commit crimes in the first place. But a system that would work when being gamed by people desperate not to go to jail?
Alternatively, is there a relevant moral principle that people can follow unilaterally that would make the world a better place (other than deontologically)? If we tell a defendant not to hire a lawyer, or a lawyer not to argue as well as they can (while keeping to lawyer ethics), or the jury not to listen to the lawyers—then the side that doesn’t cooperate will win the trial, or the jury will ignore important claims, and justice won’t be better served on average.
Are you sure? I’ve met a lot of people (“average people”, not rationalists) who take the view of “yeah, he can talk real impressively, but it’s all bullshit, no doubt”. Many people like “simple talk”, i.e. speech that simply lays out facts, and are suspicious of impressive/skillful rhetoric.
I’ve met fewer people like that, but then I’m not a native English speaker, so not all of the speechifying I’m exposed to is in English.
It sounds like the thing being described is in part a desire for the speaker to talk in a particular dialect and style, associated with a social class or background, with the appropriate choice of acrolect/mesolect/basilect, etc.
Do you think such people are unaware that the “plain” speakers they like still harness rhetoric, just one tailored to that audience? Good plain speech still needs to be concise, address the right points, have good body language, good delivery (e.g. not stutter or repeat yourself), and in the end say things that the audience will like as speech as well as on the object level. An untrained speaker will rarely carry an audience, however plainly they speak.
This seems to depend a lot on social context.
Lawyers are the quintessential speakers-for-hire who apply rhetoric to mercenary causes. Yet lawyers are accepted and high-status in most parts of society. In play, debate clubs are often popular; and at a recent LW meetup we played an Ideological Turing Test game, where we had to convince each other of positions we didn’t always hold.
Lawyers don’t hide the fact they’re biased and mercenary in court. Maybe that helps them disassociate enough from the practice that people don’t feel uncomfortable with them in a personal setting. And yet, most people are not bothered by the idea that the justice system runs in part on rhetoric persuasiveness, or that the wider political system runs in part on politicians convincing voters of things. People compliment politicians on their public speaking skills without thinking “dark arts!” every time.
In general, the average person’s reaction to “wow what a skilled orator” isn’t “therefore I won’t listen to them, persuasiveness is orthogonal to truth”. How do you reconcile this with your analysis of “betraying cooperative norms”, which people are usually good at enforcing?
Note: I’m not well familiar with modern lawyers and exactly how important rhetoric is to them. In classical antiquity it was extremely important; that is relevant insofar as it partially explains why rhetoric is present in today’s classical liberal education.
Unlike the regular trolley problem, this one is similar to moral choices made by many people every day.
As you walk down the street, you see a man point a gun at another. Do you interpose your body to save the stranger’s life at the cost of your own?
Down another street, a man points a gun at you. There’s a bystander nearby. Do you hide behind this human shield to save your life?
Some people will view this question differently from the trolley one because the attacker is a human and can be assigned blame. Substitute a grisly bear that is attacking a group of campers. Do you shield another with your body, or conversely pull another’s body in front of yours?
I agree, and I didn’t mean to imply otherwise.
The correlation between moral objectivism and interventionism is probably true, but I think it’s historically contingent, and not a logical consequence of objectivism. Whether or not I think of my morality as objective (universal) or subjective (a property of myself), that’s orthogonal to what I actually think is moral.
I’m a moral relativist. My morality is that torture and murder are wrong and I am justified and, indeed, sometimes enjoined to use force to stop them. I don’t think this is an uncommon stand.
Other people are moral objectivists, but their actual morals may tell them to leave others alone except in self-defense.
I haven’t listened to the debate (I’d read it if it was transcribed), but I want to object to a part of your post on a meta level, namely the part where you say:
To me, he is very far from a model for a rationalist
Being able to effectively convince people, to reliably influence their behavior, is perhaps the biggest general-purpose power a human can have. Don’t dismiss an effective arguer as “not rationalist”. On the contrary, acknowledge them as a scary rationalist more powerful than you are.
The word “rationalist” means something fairly narrow. We shouldn’t make it into an applause light, a near synonym of “people we like and admire and are allied with”. Being reliably effective, on the other hand, is a near synonym of being rational(ist).
If Adams employed “dark arts” in his debate, the only thing that necessarily means is that he wasn’t engaged in an honest effort to discover the truth. But that’s not news—it was a public debate staged in order to convince the audience! So Adams used a time-honored technique of achieving this goal—how very rational of him. At least, it’s rational if he succeeded, and I assume you think he did succeed in convincing some of the audience, otherwise you wouldn’t bother to post a denunciation.
Similarly, the name “Dark Arts” is misleading. They are (if I may channel Professor Quirrell for a moment) extremely powerful Arts everyone should cultivate if they can, and use where appropriate: not when honestly conversing with a fellow rationalist to discover the truth, but when aiming to convince people who are not themselves trained in rationality, and who (in your estimation) will not come by their beliefs rationally, whether or not they end up believing the truth.
This is a near cousin of politics (in the social sense, not the government sense). Politics is a mind-killer and it’s important to keep politics-free spaces for various purposes including the pursuit of truth. But we should not say “rationalists should not engage in politics”, any more than “rationalists should never try to convince non-rationalists of anything”.
ETA: I’m not claiming Adam is a rationalist or is good at being a rationalist; I’m not familiar enough with him to tell. I’m only claiming that the fact he is or tries to be a good persuader in a debate and uses Dark Arts isn’t evidence that he isn’t.
I feel the briefness of history is inseparable from its speed of change. Once the agricultural revolution got started, technology kept progressing and we got where we are today quite quickly—and that’s despite several continental-scale collapses of civilization. So It’s not very surprising that we are now contemplating various X-risks: to an external observer, humanity is a very brief phenomenon going back, and so it’s likely to be brief going forward as well. Understanding this on an intuitive level helps when thinking about the Fermi paradox or the Doomsday Argument.
I think you’re mostly right about that, but not entirely. The two realms are not so clearly separated. There are social hangouts on the Internet. There are social hangouts, of both kinds, where people talk shop. There are programming blogs and forums where social communities emerge. And social capital and professional reputation feed into one another.
So that’s the real role of the expert here
I work in the data science industry—as a programmer, not a data scientist or statistician. From my general understanding of the field what you’re describing is a broadly accepted assumption. But I might be misled by the fact that the company I work for bases its product on this assumption, so I’m not sure if you’re just describing this thing from another angle or if there’s a different point that I’m missing here or if, in fact, many people spend too much effort trying to hand-tune models.
The data scientists I work with make predictive models in two stages. The first one is to invent (or choose) “features”, which include natural variables from the original dataset, functions acting on one or more variables, or supplementary datasets that they think are relevant. The data scientist applies their understanding of statistics as well as domain knowledge to tell the computer which things to look for and which are clearly false positives to be ignored. And the second stage is to build the actual models using mostly standard algorithms like Random Forest or XGBoost or whatnot, where the data scientist might tweak arguments but the underlying algorithm is generally given and doesn’t allow for as much user choice.
A common toy example is the Titanic dataset. This is a list of passengers on the Titanic, with variables like age, name, ticket class, etc.. The task is to build a model that predicts which ones survived when the ship sank. A data scientist would mostly work on feature engineering, e.g. introducing a variable that deduces passenger’s sex from their name, and focus less on model tuning, e.g. determining the exact weight that should be given to the feature in the model (women and children had much higher rates of survival).
In a more serious example, a data scientist might work on figuring out which generic datasets are relevant at all. Suppose you’re trying to predict where to best open a new Starbucks branch. Should you look at the locations of competing coffee shops? Noise from nearby construction? Public transit stops or parking lots? Nearby tourist attractions or campuses or who knows what else? You can’t really afford to look at everything, it would both take too long (and maybe cost too much) and risk false positives. A good domain expert is the one who generates the best hypotheses. But to actually test those hypotheses, you use standard algorithms to build predictive models, and if a simple linear model works, that’s a good thing—it shows your chosen features were really powerful predictors.
Even computer programmers who spent the majority of their working output working alone can benefit a lot from having good connections when it comes to finding good jobs.
People skills have great value for programmers, and finding jobs is a very small part of it. I write this from personal experience.
Programmers are still people. The amount of great software any one person can write in their lifetime is very limited. Teaching or convincing others (from coworkers to the rest of the world) to agree with you on what makes software great, to write great software themselves, and to use it, are the greatest force multipliers any programmer can have, just like in most other fields.
Sometimes there are exceptions; one may invent a new algorithm or write some new software that everyone agrees is great. But most of the time you have to convince people—not just less-technical managers making purchasing decisions, but other programmers who don’t think that global mutable state is a problem, really, it worked fine in my grandfather’s time and it’s good enough for me.
This post is well written and not over-long. If the concepts it describes are unfamiliar to you, it is a well written introduction. If you’re already familiar with them, you can skim it quickly for a warm feeling of validation.
I think the post would be even better with a short introduction describing its topic and scope, but I’m aware that other people have different preferences. In particular:
There are more than two ‘cultures’ or styles of discussion, perhaps many more. The post calls this out towards the end (apparently this is new in v2).
The post gives two real examples of Combat Culture, and only one made-up scenario of Nurture Culture. It does not attempt to ground the discussion in anything quantitative—how common these cultures are, what they correlate with, how to recognize or test for them, how gradually they may shade into each other or into something else altogether.
I don’t want to frame these as shortcomings; the post is still useful and interesting without them!