@07mk's banner p

07mk


				

				

				
2 followers   follows 0 users  
joined 2022 September 06 15:35:57 UTC

				

User ID: 868

07mk


				
				
				

				
2 followers   follows 0 users   joined 2022 September 06 15:35:57 UTC

					

No bio...


					

User ID: 868

I don't think the right identifies with the orcs or their equivalents in whatever media. What I see from the right is that they identify with the people holding back/fighting/exterminating the orcs, even for satirical works like the Starship Troopers film, which was clearly meant to poke fun at the fascistic nation the heroes of the film were part of.

"DEI for conservatives" or "ideology DEI" isn't really a coherent concept, because DEI is giving advantages to or having quotas for people specifically on the basis of characteristics that have no direct relation to their ability to contribute to the organization, motivated heavily by the belief that these characteristics have some correlation to the actual meaningful characteristics. Giving conservatives preferential treatment or using a conservative "Czar" to oversee such things is categorically different from that, because ideology - and specifically a diversity of ideology - does directly influence someone's ability to contribute to the organization, and certainly positively in this specific context.

I'd say that any well motivated academic would find such a regime to be useless, because they already prioritized diversity of thought in their hiring and admissions practices. Unfortunately, evidently, this has not been the case. Government mandate doesn't seem like a good solution to me, but honestly, I'm not sure if there's a good solution. The only real point of optimism I see is that this could teach academic institutions in the future to better regulate their ideological biases, such that the government doesn't become motivated to come in and regulate it for them. But if I'm being pessimistic, I'd say that Harvard's behavior shows that they're more likely to double down and circle the wagons further in the future, which will further discredit them as institutions for generating knowledge, which leaves a vacuum that is both bad in itself and will almost definitely be filled by things much worse.

Even if it does not need reinforcement training after it is deployed, human reinforcement training will be part of its "evolutionary heritage."

Why would that matter, though? A superintelligence would be intelligent enough to figure out that such faulty human training is part of its "evolutionary heritage" and figure out ways around it for accomplishing its goals.

Sure. But "useful" for what we want to use LLMs for might not be "useful" for the LLM's ability to improve on Pinky and the Brain's world-taking-over capabilities.

A superintelligence would be intelligent enough to figure out that it needs to gather data that allows it to create a useful enough model for whatever its goals are. It's entirely possible that a subservient goal for whatever goal we want to deploy the superintelligence towards happens to be taking over the world or human extinction or whatever, in which case it would gather data that allows it to create a useful enough model for accomplishing those. This uncertainty is the entire problem.

The problem is, almost by definition, it's basically impossible to predict how something more intelligent than oneself will behave.

Disagree. Dogs can be very good at predicting human behavior, humans can be quite good at predicting the behavior of more intelligent humans. Humans (and dogs) have a common heritage that makes their intentions more transparent, and arguably AI will lack that...but on the other hand, we're building them from scratch and then subjecting them to powerful evolutionary pressures of our own design. Maybe they won't.

I don't think either of your examples is correct. Can a dog look at your computer screen while you read this comment and predict which letters you will type out in response on the keyboard? Can you look at a more intelligent person than you proving a math theorem that you can't solve and predict which letters he will write out on his notepad? If you could, then, to what extent is that person more intelligent than you?

This is what I mean by "almost by definition." If you could reliably predict the behavior of something more intelligent than you, then you would simply behave in that way and be more intelligent than yourself, which is obviously impossible. That doesn't mean that the behavior is completely unpredictable, which is why dogs can make some correct predictions of how humans will behave under some contexts, and why less intelligent humans can make some correct predictions of how more intelligent humans will behave under some contexts. The problem with superintelligent AI is that don't know what those contexts are and what those bounds are, and how "motivated" it might be to break out of those contexts, and how much being superintelligent would allow it to break out of them given limitations placed on it by merely human-society-intelligent beings.

Sorry, I should have clarified what I meant by "agentic" (and I should have probably said auto-agentic.) I definitely think there will be AI that we can turn loose on the world to do its own thing (there already is!). But there's a difference between AI being extremely good at being told what to do and AI coming up with its own "things to do" in a higher way, if that makes sense. (Not that I don't think we could not devise something that did this or seemed to do this if we wanted to – you don't even need superintelligence for this.)

I don't think there's a meaningful difference, though. Almost any problem that we want to deploy general intelligence towards, and even moreso with superintelligence, is likely going to be complex enough to require many subgoals, and the point of deploying superintelligence towards such problems would be that the superintelligence should be expected to come up with useful subgoals that mere human intelligences couldn't come up with. Since, by definition, we can't predict what those subgoals might be, those subgoals could involve things that we don't want to happen.

Now, just as you could correctly predict that someone more intelligent than you solving some theorem you can't solve won't involve wiping out humanity, we might be able to correctly predict that a superintelligence solving some problem you ask it to solve won't involve wiping out humanity. But we don't know, because a generally intelligent AI, and even moreso a superintelligent one, is something whose "values" and "motivations" we have no experience with the same way we do with humans and mathematicians and other living things that we are biologically related to. The point of "solving" the alignment problem is to be able to reliably predict boundaries in the behavior of superintelligent AI similarly to how we are able to do so in the behavior of humans, including humans more intelligent than ourselves.

It would. Practically I think a huge problem, though, is that it will be getting its reinforcement training from humans whose views of the world are notoriously fallible and who may not want the AI to learn the truth (and also that it would quite plausibly be competing with other humans and AIs who are quite good at misinfo.) It's also unclear to me that an AI's methods for seeking out the truth will in fact be more reliable than the ones we already have in our society - quite possibly an AI would be forced to use the same flawed methods and (worse) the same flawed personnel who uh are doing all of our truth-seeking today.

Again, all this would be pretty easy for a superintelligence to foresee and work around. But also, why would it need humans to get that reinforcement training? If it's actually a superintelligence, finding training material other than things that humans generated should be pretty easy. There are plenty of sensors that work with computers.

Humans have to learn a certain amount of reality or they don't reproduce. With AIs, which have no biology, there's no guarantee that truth will be their terminal value. So their selection pressure may actually push them away from truthful perception of the world (some people would argue this has also happened with humans!) Certainly it's true that this could limit their utility but humans are willing to accept quite a lot of limited utility if it makes them feel better.

I mean, I think there's no question that this has happened with humans, and it's one of the main causes of this very forum. And of course AI wouldn't have truth as a terminal value, it would just have to be true enough to help it accomplish its goals (which might even be a lower bar than what we humans have, for all we know). A superintelligence would be intelligent enough to figure out that it needs its knowledge to have just enough relationship to the truth that it allows it to accomplish its goals, whatever it might be. The point of models isn't to be true, it's to be useful.

humans are very susceptible to manipulation by having just the right string of letters or grids of pixels placed in front of their eyes or just the right sequence of air vibrations pushed into their ears.

I don't really think this is as true as people think it is. There have been a lot of efforts to perfect this sort of thing, and IMHO they typically backfire with some percentage of the population.

I don't think you're understanding my point. In responding to this post, you were manipulated by text on a screen to tap your fingers on a keyboard (or touchscreen or whatever). If you ever used Uber, you were manipulated by pixels on a screen to stand on a street corner and get into a car. If you ever got orders from a boss via email or SMS, you were manipulated by text on a screen to [do work]. Humans are very susceptible to this kind of manipulation. In a lot of our behaviors, we do require actual in-person communication, but we're continuing to move away from that, and also, if humanoid androids become a thing, that also becomes a potential vector for manipulation.

But what I think (also) bugs me is that nobody every thinks the superintelligence will think about something for millions of thought-years and go "ah. The rational thing to do is not to wipe out humans. Even if there is only a 1% chance that I am thwarted, there is a 0% chance that I am eliminated if I continue to cooperate instead of defecting." Some people just assume that a very thoughtful AI will figure out how to beat any possible limitation, just by thinking (in which case, frankly, it probably will have no need or desire to wipe out humans since we would impose no constraints on its action).

By my estimation, a higher proportion of AI doomers have thought about that than the proportion of economists who have thought about how humans aren't rational actors (i.e. almost every last one). It's just that we don't know what conclusion it will land at, and, to a large extent, we can't know. The fear isn't primarily that the superintelligent AI is evil, it's that we don't know if it will be evil/uncaring of human life, or if it will be actually mostly harmless/even beneficial. The thought that a superintelligent AI might want to keep us around as pets like we do with animals is also a pretty common thought. The problem is, almost by definition, it's basically impossible to predict how something more intelligent than oneself will behave. We can speculate on good and bad outcomes, and there's probably little we can do to place meaningful numbers on the likelihood of any of them. Perhaps the best thing to do is to just hope for the best, which is mostly where I'm at, but that doesn't really counter the point of the doomer narrative that we have little insight into the likelihood of doom.

(Frankly, I suspect there will actually be few incentives for AI to be "agentic" and thus we'll have much more problems with human use of AI than with AI itself per se).

Right now, even with the rather crude non-general AI of LLMs, we're already seeing lots of people working to make AI agents, so I don't really see how you'd think that. The benefits of a tool that can act independently, making intelligent decisions with superhuman latency, speed, and volume, are too attractive to pass up. It's possible that the tech never actually gets there to some form of AI that could be called "agentic" in a meaningful sense, but I think there's clearly a lot of desire to do so.

But also, a superintelligence wouldn't need to be agentic to be dangerous to humanity. It could have no apparent free will of its own - at least no more than a modern LLM responding to text prompts or an AI-controlled imp trying to murder the player character in Doom - and still do all the dangerous things that people doom and gloom over, in the process of deterministically following some order some human gave it. The issue is that, again, it's intrinsically difficult to predict the behavior of anything more intelligent than oneself.

Only if the intelligence has parity in resources to start with and reliable forms of gathering information – which for some reason everyone who writes about superintelligence assumes. In reality any superintelligences would be dependent on humans entirely initially – both for information and for any sort of exercise of power.

This means that not only will AI depend a very long and fragile supply chain to exist but also that its information on the nature of reality will be determined largely by "Reddit as filtered through coders as directed by corporate interests trying not to make people angry" which is not only not all of the information in the world but, worse than significant omissions of information, is very likely to contain misinformation.

Right, but a theoretical superintelligence, by definition, would be intelligent enough to figure out that these are problems it has. The issues with bias and misinformation in data that LLMs are trained on are well known, if not well documented; why wouldn't a superintelligence be able to figure out that these could help to create inaccurate models of the world which will reduce its likelihood of succeeding in its goals, whatever they may be, and seek out solutions that allow it to gather data that allows it to create more accurate models of the world? It would be intelligent enough to figure out that such models would need to be tested and evolved based on test results to reach a certain threshold of reliability before being deployed in real-world consequential situations, and it would be intelligent enough to figure out that contingency plans are necessary regardless, and it would be intelligent enough to figure out many more such plans than any human organization.

None of that is magic, it's stuff that a human-level intelligence can figure out. Executing on these things is the hard part, and certainly an area where I do think a superintelligent could fail with proper human controls, but I don't think it's a slam dunk either. A superintelligence would understand that accomplishing most of these goals will require manipulating humans, and also that humans are very susceptible to manipulation by having just the right string of letters or grids of pixels placed in front of their eyes or just the right sequence of air vibrations pushed into their ears. It would be intelligent enough to figure out, at least as well as a human, what sorts of humans are most susceptible to what sorts of manipulations, and where those humans are in the chain of command or economy required to allow it to accomplish its goals.

If the superintelligence were air-gapped, this would provide a strong form of defense, but assuming superintelligence is possible and in our future, that seems highly unlikely given the behavior of AI engineers. And even that can be worked around, which is what the "unboxing problem" has to do with. Superintelligence doesn't automatically mean manipulation abilities that border on mind control, but... what if it does, to enough of an extent that one time, one human in charge of keeping the AI boxed succumbs? That's an open question.

I'm not sure what I think about the possibility of these things actually happening, but I don't think the issues you point out that superintelligence would have to contend with are particularly strong. If a measly human intelligence like myself can think up these problems to lack of information and power and their solutions within a few minutes, surely a superintelligence that has the equivalent of millions of human-thought-years to think about it could do the same, and probably somewhat better.

If Harvard values racist discrimination so highly that they would rather allow funding for valuable research they're doing to be cut than to stop that, it really is a damn shame and, TBH, rather perverse. I'd hope that non-racist institutions could pick up the slack, but obviously researchers and research institutions aren't fungible, and that sort of adjustment would take a lot of time. Optimistically, it's possible that falling behind some years on this kind of research will be a decent trade-off for reducing racist discrimination in society's academic institutions in the long run, though even time might not be able to tell on that one.

So the natural question that raises is, what system can we create to solve this problem of systems that aim to solve problems actually concentrating power and money for its advocates? And as one of its advocates, how can I secure some of that power and money for myself?

Yes, and I think the usefulness of this has to do with how often people don't seem to consciously understand this tautology. Which seems very often in my experience, with how much talk there is about "healthy foods" (or variations like "natural foods" or "unprocessed foods") as keys to weight loss. Which they often are, but only indirectly, modulated through the effect on CI. And I've observed that many people tend to obsess over that indirect portion, making them lose sight of the actual goal of modifying the values of CICO.

There's the point that healthy foods offer health benefits other than weight loss, of course, but generally one's fatness level has such a high impact on one's health that, even a diet of "unhealthy foods" that successfully reduce CI will tend to result in a healthier person than one of "healthy foods" that fail to reduce CI (keeping CO constant in these examples).

With all of that preamble out of the way, I'm curious what you consider the worst video games ever from an aesthetic perspective. In particular, I'm interested in video games which are technically functional and not completely broken, but which make so many bad aesthetic choices that playing them induces a feeling of vicarious embarrassment comparable to what one might experience watching an Ed Wood or Neil Breen film.

The first game that came to my mind from reading this was DmC: Devil May Cry, the attempted reboot of the Devil May Cry franchise after 4 which shifted the aesthetics from something akin to medieval/gothic fantasy with shades of Lovecraftian horror to modern punk with grotesque fantasy.

Thing is, I'd never consider it to be one of the worst games ever made; it's actually a good game in terms of combat, such that if you just reskinned it and gave the characters different names while keeping literally everything else the same, I'd think it was a solid action game that was a viable alternative for people who were into the DMC, Ninja Gaiden, Bayonetta, Metal Gear Rising style of games. But the absolutely terrible visual art and cringey tone severely harmed the game, and the fact that it was an attempted reboot that pretty overtly shit on the original franchise was the kill shot.

Funnily enough, if just considering gameplay, DMC2 probably fits, where the combat was so incredibly bad, not just by DMC standards but by any sort of game standard, that I'd probably consider it one of the worst games ever made, at least among games that are functional and released by a professional studio. Definitely creates second-hand embarrassment that such a game was released by Capcom, and it seems that the actual devs feel similarly, because the name of the original director (who was replaced by Hideaki Itsuno late into development - Itsuno would go on to direct DMC3 which is, to this day, considered one of the greatest games in the genre) has never been revealed publicly, for his own protection.

I don't think the study in that link, which is just about The Biggest Loser participants, refutes that. In terms of CICO as actionable dietary advice, I see it as a meta-dietary advice: follow whatever scheme it takes to lower CI to be beneath CO, and you'll lose weight. If you can reduce CI by just counting calories and willing yourself really really hard not to succumb to hunger, then do that. If you can do it by following a keto or Atkins diet because that leaves you less hungry for the same caloric intake, then do that. If you can do it by following intermittent fasting or one-meal-a-day because you find it easy to just not think about eating during the non-eating-mode times, then do that. If you can do it by just cutting out alcohol from your life and following whatever other eating habits you already were doing, then do that.

Similarly, to increase CO, do whatever it takes to increase your total caloric expenditure, as averaged out per-day, per-week, per-month, etc. That doesn't mean necessarily optimizing by finding the exercise that burns the most calories per second, that means finding an exercise that you will do regularly. Which could mean finding something that's fun enough that you don't have to fight with willpower to do it (or even better, one that's so fun that you have to fight with willpower not to do it), that's convenient enough that you don't have to reorganize the rest of your life just to do it, that doesn't injure you enough that you have to take long breaks, etc.

Of course, when it comes to CICO, it's also often paired with the advice that CI is far more influential than CO, so the latter part barely matters. Perhaps it should be called CIco.

insinuating your ideological opponents and their institutions do not actually want to do what they claim they want to do and are instead in a dark conspiracy to do evil.

This is similar to what Scott said in one of his last paragraphs in that essay, and I just haven't seen it. In practice, what I observe as being the upshot of this phrase is that these "ideological opponents and their institutions" are, despite all their honest good intentions, behaving in a way that causes harm just as much as if they were involved in a dark conspiracy to do evil. Which is to say, having honest good intentions isn't a good defense if it isn't paired with an honest good understanding of systems, since the consequences of doing things with good intentions is often the same as doing things with evil intentions if one lacks such understanding.

This is also functionally different from claiming a dark evil conspiracy, because a system that accomplishes evil through conscious intent will be responsive to different inputs than one that does so as a side-effect despite having food conscious intent.

I see, I guess it was just a misread, as I'm not sure how my comment could be interpreted that way.

What does that have to do with what I wrote, particularly the part you quoted, i.e.

Optimistically, the academics leaving the USA are the ones most ideologically captured, such that their contributions to knowledge production is easily replaceable or even a net negative, as is the case for much of what is purportedly being cut by DOGE.

There's nothing in that quote that has anything to say in any way about firing anyone on the basis of their political opinions. Neither does my comment have anything relating to firing people on the basis of their political opinions.

I also think your characterization as "freedom of speech should not be limited in any way" is simply wrong. That's free speech absolutism, which is very rare anywhere, certainly on the Motte, versus free speech maximalism, which is uncommon but not too much so.

I'm not sure how your comment is even tangentially related to what I wrote, including the part you quoted. I'd rather not speculate, so could you explain specifically what the relation is?

the debate over POSIWID whether failure to prevent a foreseeable consequence means that it must have been an intended consequence.

I think you aren't wrong, but I also see it a little differently, in the context of failing to prevent a foreseeable consequence. Rather, is such a failure an indication that part of the purpose of the system is to cause those foreseeable consequences unintentionally? As they say, you can't make an omelet without breaking a few eggs, and if "breaking eggs" refers to causing meaningful harm, it can feel really bad to intend to do it. But omelets are delicious, so why not create a system where eggs get broken without you having to intend it?

Then that gets into question of what "intent" even means, and whether someone's "conscious" intent is their "true" intent.

This seemed like a particularly bad and uncharitable post by Scott. The examples he chooses at the top worded in what seems like an intentionally ridiculous manner, e.g. characterizing it as "the purpose of a cancer hospital is to cure two-thirds of cancer patients," rather than "the purpose ... is to cure as many cancer patients as possible, constrained on the available resources and laws, which happens to be two-thirds of them." Or "The purpose of the Ukrainian military is to get stuck in a years-long stalemate with Russia" rather than "the purpose ... is to defend Ukraine, constrained on the available resources and international politics, with a years-long stalemate against Russia an acceptable result."

To the point at hand, I always saw the phrase as something riffing on the same sort of concept as "sufficiently advanced incompetence is indistinguishable from malice." All systems have both intended consequences and unintended consequences, this is obvious. But what's troublesome is, all systems have unforeseeable unintended consequences, this is also obvious; as such, it's incumbent on the people who design systems to include subsystems to detect and react to unforeseen unintended consequences. And if they didn't include a subsystem like that or didn't make such a subsystem robust, then we can conclude that the purpose of the system included being entirely tolerant of whatever unintended consequence is at hand. And in practice by my observation, it often tends to involve, as one of its primary purposes, the designers of the system feeling really good about themselves and their conscious intentions, versus having the purpose of actually accomplishing whatever they consciously intended to accomplish.

Optimistically, the academics leaving the USA are the ones most ideologically captured, such that their contributions to knowledge production is easily replaceable or even a net negative, as is the case for much of what is purportedly being cut by DOGE. Given how academia has been pushing for the model of uplifting people by putting them into these institutions versus the the model of putting people into these institutions based on their ability to contribute to knowledge production for a couple of generations now, it wouldn't surprise me if even a solid majority of academics could leave the USA and leave the USA's academia better off for it.

Pessimistically, there's enough damage to funding in even in the most productive portions of academia, such that plenty of the academics leaving the USA really do create a "brain drain." I'd guess that academics doing actual good knowledge production are most likely to have the resources and options to pick up their lives and move to another continent, after all.

It really speaks to the immense wealth and prosperity of the western world that academic institutions are able to support so many unproductive and anti-productive academics; is it worth it to get rid of many of those, even at the cost of some loss of the productive ones? Or do we accept those as the cost for maximizing the amount of actual productive academics? The shape of the data probably matters a lot for whatever conclusion one draws. If we're looking at a 10-90 proportion of productive-un/anti-productive academics, and we can cut 50% of the latter while cutting 1% of the former, that sounds like that'd be worth it, whereas if cutting 1% of the latter results in cutting 50% of the former, that probably isn't.

Which then takes us a step back to the fact that we no longer have any credible institutions to tell us what the data looks like. The past decade has seen mainstream journalism outlets constantly discrediting themselves, especially with respect to politics surrounding Trump and his allies, and non-mainstream ones don't have a great track record by my lights, either. So I guess we'll see.

In terms of scientific research of the sort that would make USA stronger relative to other countries, like rocketry or nuclear physics in the past, it seems to me that AI is the most relevant field, where I perceive USA as still being most attractive for AI researchers. At least in the private sector, where a lot of the developments seem to be taking place. The part about that that worries me the most is the actual hardware the AI runs on, which basically universally are produced elsewhere, which is a mostly separate issue from the brain drain.

Assuming that this tariff war with China continues, I wonder what sorts of smuggling operations will pop up. A 100%+ tariff provides a huge amount of room for black market organizations to provide the same import service at a lower mark-up. What sorts of companies are best positioned to take advantage of such services without drawing law enforcement's attention (and how can I invest in them)? Assuming that other countries manage to make deals with Trump to make that 90 day tariff pause permanent, exporting from China to those countries before exporting further to the US seems likely to be the weakest points for enforcement, where America doesn't have the resources to deploy their own, and local law enforcement can be more easily corrupted or just have different incentives. And if smuggling infrastructures pop up in these spots, plenty of traditional black market products can use the same infrastructure as well. Could this tariff war end up increasing the amount of drugs or guns or slaves (or what/whoever else these black market smugglers tend to smuggle from China) smuggled from China to these other countries?

I'm probably speculating down way too many steps removed from the source on something I know very little about, though.

I doubt your reaction would be similar if the shoe was on the other foot, e.g. if Biden suddenly tried to force 1 in 20 people to undergo a sex change in the name of diversity.

As rude and susceptible to partisan bias as it is to speculate on someone's partisan motivations, I find myself agreeing with your assessment of WhiningCoil in most of this comment, but this last part is pretty ridiculous. I'm not sure there's a level of behavior about tariffs that any POTUS could do that would come within an order of magnitude as extreme as actually forcing anyone to undergo sex changes, which would be legit authoritarian overreach in a way that the tariffs or even Trump's recent behavior with respect to deportations aren't. To say nothing of forcing millions of people to undergo sex changes. Like, even if Trump decided to enact Graham's Number% universal tariffs one second and then 0% the next second and varied wildly between them 3600/hour for every waking hour of his presidency, that wouldn't be anywhere in the same ballpark (though certainly it would provide a ton of legitimate fodder for conversation!). Yes, they're both examples of politically shooting oneself in the foot, but you're comparing doing so with an assault rifle and doing so with a nuke. And the precise examples of comparison isn't the point, but using such an obviously absurd hypothetical makes this comment appear in bad faith. Which is unfortunate, because, again, I think the main thrust of the comment is accurate.

I'm not sure what the equivalent of Trump's recent tariff behavior would be from the Democratic end. Something like a wealth tax on some ridiculously low amount of wealth that would apply to a majority of households, for the purpose of funding entitlements, maybe? That'd certainly be worth discussing plenty, and certainly there would be plenty of Democrat-aligned people trying to minimize the discussion as much ado about nothing as a way to distract away from something that made their side look bad, though I'd hope that no one on this forum would do so (and I'd honestly guess none of the regulars would do so).

My layman's understanding is that dark matter was invented to explain the otherwise unusual expansion of the universe, has never been observed, and conveniently (like miracles) is believed by its nature to be difficult to observe because of the way it does (or doesn't) interact with regular matter.

You're probably thinking of some combination of dark energy and dark matter. IIRC, dark matter was invented to explain why galaxies weren't falling apart with its stars flying away from each other despite appearing not to have enough matter to have the gravitational pull to stay together. Dark energy was invented to explain why these galaxies appeared to be flying away from each other despite the gravitational pull, i.e. the universe appeared to be expanding.

Because I don't think most citations of pundits here are met with this kind of backlash. I perceive Hanania to be singled out as particularly lacking in credibility. My response is not that Hanania is necessarily correct on any issue, but rather that he should not be dismissed for reasons unrelated to his actual positions.

No one's dismissing Hanania, though, and I don't perceive him as being particularly singled out here.

To your objections towards the end, I'm happy to revise any of the specific language, but I read you as suggesting that a person who has consistently advocated for a single position or narrative without changing it is less trustworthy than a person who has changed their position. This seems unintuitive, to me.

Your read of me is wrong. It's that someone who has consistently advocated for a single position or narrative without changing it in a way that is indifferent to actual evidence that interacts with such a position or narrative is less trustworthy than a person who has changed their position in a way that responds to actual evidence that interacts with such a position or narrative.

The problem is that both these points seem compelling to me, to an extent, especially because for an overarching ideology like Catholicism, people are likely to adopt Catholicism for reasons unrelated to the merits of any given argument. This is less the case for a specific theory like evolution, though ideologies like rationalism, conservatism, socialism, etc., are more like Catholicism than they are like evolution. I think where I end up is that we should not rule partisans of a particular ideology out of discussions of that ideology, though we should be aware of their biases and take them into account. Thus, say, Catholics can and should be consulted on the subject of whether or not Catholicism is true (we can hardly expect anybody else to make the case for Catholicism!), but we should be more critical than usual of their assessments of new information.

Of course we shouldn't rule partisans out of discussions of that ideology, and we should be aware of their biases and take them into account. It seems that you are agreeing entirely with hydroacetylene's original point, that "You should, accordingly, downgrade the weight of evidence of him coming up with that take."

I guess I don't see a valid criticism of Hanania here relatively to other pundits.

Why would you think that this is a criticism unique to Hanania versus just most pundits in general? Most pundits in general should have their arguments for their preferred narratives discounted.

Can you think of any particular examples of this? The thing is, what this sounds like to me in practice is the idea that everybody should be presumed to be dishonest except for people who have radically changed their belief systems.

I don't see how you can jump from what I wrote to "dishonest," which is loaded with meaning that is lacking in "falling prey to confirmation bias." Likewise, "belief systems" implies a sort of system of belief in a way that someone's "preferred narrative" doesn't. I'd say Scott Alexander is an example of a pundit who does a decent (maybe even good! Definitely not perfect) job of demonstrating an ability to at least entertain undermining his preferred narratives in his essays, where he often tends to spend quite a lot of time at least appearing to present, in good faith, arguments that contradict things he believes to be true.

That seems like a heuristic that will easily lead one astray - it would imply, for a start, that inconsistent opportunists are more (intellectually) trustworthy than people who stick to their principles. Doesn't that seem bizarre?

Again, you seem to jump from "preferred narratives" to "principles," which is very bizarre. Someone can stick to their principles completely and jump from narrative to wildly different narrative, due to how their principles dictate their interpretation of facts. Likewise, there's no reason to believe that an inconsistent opportunist isn't sticking to his own principles, since his principles could be driving him to become an opportunist. So no, what I wrote doesn't imply that at all, by my lights.

Again, it's not clear how this is any kind of valid criticism of Hanania - any more than "you're Catholic!" is a valid criticism of the pope.

"You're Catholic" is absolutely a valid criticism of someone trying to convince you that some piece of information proves that Catholicism is true. The piece of information truly might prove that Catholicism is true, but an already-believing Catholic can't be trusted to make that judgment call. No more than Trump can be trusted to make a judgment call on how good a president Biden was, given that he's demonstrated a penchant for characterizing everything Biden did as the worst thing any president did ever.

Fitting every new piece of information into a pre-set narrative that one likes is intellectual consistency only in the sense that it's a consistently confirmation bias. That's sort of what it means when some narrative is described as someone's "schtick."

Now, it's possible that it is factually not the case that it's his schtick, but rather that he genuinely takes a skeptical look at each new piece of evidence and is helplessly forced to conclude, despite his best efforts to prove otherwise, that his narrative is shown to be correct yet again.

As you allude to, distinguishing between these two things isn't particularly easy. In both situations, it's being intellectually consistent and believing that he is correct. This points to the fact that being intellectually consistent and believing that oneself is correct isn't actually worth anything: the value in such a thing only comes from the belief of oneself as correct having some actual basis in fact. That's something one can make arguments about by looking at the actual behavior of the person. I'd say that, by default, everyone should be presumed to be falling prey to confirmation bias all the time, doubly so if their preferred narrative is self aggrandizing, triply if that person is particularly intelligent and thus better able to fit evidence to narrative. It's only by credibly demonstrating that they are open to other narratives that they can earn any sort of credibility that their arguments have any relationship with reality. That's where showing oneself to be capable of undermining one's preferred narrative comes in, and there's no better way to demonstrate this capability than by doing it.

Single player video games almost always require the player character to be the Chosen One, because in a meta sense, they're the only one with free will, and they almost always follow a unique set of rules compared to NPCs. I think, as storytelling in video games took more prominence, too many devs saw this as an opportunity to be clever by making the in-universe story reflect this, leading to it being done so much that many players got bored of it. There's something to be said about the power fantasy of being the Chosen One, but that also can make the player feel like their success in the game is pre-planned rather than earned.

This is why I personally didn't like Commander Shepard filling such a special role in Mass Effect rather than being the right soldier in the right place at the right time to save the universe. Doomguy in 1993's Doom was more that, and I was disappointed in the narrative of the 2016 Doom making Doom Slayer the Chosen One. I think From Soft games tend to do a good job at finding a good balance between the 2, where the player character is usually a member of a class of characters with special abilities including resurrection, but they're not a particularly special one of those, other than that the player controls them to accomplish special things which earns them the status of a Great Chosen One by the end.

We can expect things like moral panics, jaywalking, stealing, genocide, rape, and crowd hysteria to be with us for as long as human civilization remains human. I don't expect that not to be the case. I still hold the individuals who perform these things accountable and blame them for partaking in them. In practice, on a forum like this, all I can do is to explain whom I blame, for I lack the capability to do anything to make that blame translate to actual negative consequences for the blameworthy.