site banner

Culture War Roundup for the week of April 10, 2023

This weekly roundup thread is intended for all culture war posts. 'Culture war' is vaguely defined, but it basically means controversial issues that fall along set tribal lines. Arguments over culture war issues generate a lot of heat and little light, and few deeply entrenched people ever change their minds. This thread is for voicing opinions and analyzing the state of the discussion while trying to optimize for light over heat.

Optimistically, we think that engaging with people you disagree with is worth your time, and so is being nice! Pessimistically, there are many dynamics that can lead discussions on Culture War topics to become unproductive. There's a human tendency to divide along tribal lines, praising your ingroup and vilifying your outgroup - and if you think you find it easy to criticize your ingroup, then it may be that your outgroup is not who you think it is. Extremists with opposing positions can feed off each other, highlighting each other's worst points to justify their own angry rhetoric, which becomes in turn a new example of bad behavior for the other side to highlight.

We would like to avoid these negative dynamics. Accordingly, we ask that you do not use this thread for waging the Culture War. Examples of waging the Culture War:

  • Shaming.

  • Attempting to 'build consensus' or enforce ideological conformity.

  • Making sweeping generalizations to vilify a group you dislike.

  • Recruiting for a cause.

  • Posting links that could be summarized as 'Boo outgroup!' Basically, if your content is 'Can you believe what Those People did this week?' then you should either refrain from posting, or do some very patient work to contextualize and/or steel-man the relevant viewpoint.

In general, you should argue to understand, not to win. This thread is not territory to be claimed by one group or another; indeed, the aim is to have many different viewpoints represented here. Thus, we also ask that you follow some guidelines:

  • Speak plainly. Avoid sarcasm and mockery. When disagreeing with someone, state your objections explicitly.

  • Be as precise and charitable as you can. Don't paraphrase unflatteringly.

  • Don't imply that someone said something they did not say, even if you think it follows from what they said.

  • Write like everyone is reading and you want them to be included in the discussion.

On an ad hoc basis, the mods will try to compile a list of the best posts/comments from the previous week, posted in Quality Contribution threads and archived at /r/TheThread. You may nominate a comment for this list by clicking on 'report' at the bottom of the post and typing 'Actually a quality contribution' as the report reason.

14
Jump in the discussion.

No email address required.

Inferential Distance part 2 of ?: Minsky's Marvelous Minutia, or why I'm bearish on GPT

This post is a continuation of / follow up to my post on Inferential distance from a month ago, inspired by the recent discussions of GPT-4 and @ymeskhout's comments on prosecutorial immunity. I also feel like this might end up turning into a series, hense the "part 2" and the question mark.

Two things that came up in that previous conversation were a) the apparent differences between thing-manipulators and symbol-manipulators. That is people's whose job, hobbies, day-to-day life revolve around manipulating objects and those whose lives revolve around manipulating symbols/feelings. And b) the question of what constitutes a "hard" social problem, and how/why thing-manipulators and symbol-manipulators seem to have such wildly diverging opinions on that question.

For a bit of context my degree is in math but entering the field as I did, later in life having already spent 12 years in another career, I tended towards the more applied/practical side of the discipline. This tendency seemed put me at odds with a lot of my instructors and fellow students, especially the "nerdier" sort. That is those who were "nerdy" even by the relative high standards of nerdiness expected from someone pursuing an advanced degree in mathematics. for whatever reason showing an interest in applications was kind of looked down upon. To be fair, I did understand where they were coming from. From a young age we're trained to admire the brilliance of guys like Pythagoras, Leibnitz, Newton, Euler, Keppler, Einstein, Et Al. Afterall, why does anyone even bother to study math if not to follow in those men's footsteps and unlock the grand fundamental truths of the universe? In contrast, while the principals of kinematics, control laws, and signal processing, may be mathematically intensive they also come across as very pedestrian. Pure math guys seem to regard them with a sort of casual disdain, the sort of thing you delegate to unpaid interns and teachers' assistants. Meanwhile truth is you can build yourself a pretty good career working on control laws and signal processing, just not in academia.

This brings us to the question of what constitutes a hard problem. If you spend enough time working in robotics or signal-processing, you'll eventually come across Moravec's Paradox. The paradox is best summed up by this xkcd comic from 2014, specifically the alt-text which reads...

In the 60s, Marvin Minsky assigned a couple of undergrads to spend the summer programming a computer to use a camera to identify objects in a scene. He figured they'd have the problem solved by the end of the summer. Half a century later, we're still working on it.

...the "paradox" being that many functions that we consider baseline, and accordingly take for granted, are in fact extremely complex and computationally intensive. Whereas much of what we might label "higher reason" is actually quite simple and requires very little in terms of memory or processing power.

It turns out that it's relatively easy to teach a computer to play chess better than a human or to come up with mathematical proofs that are both novel and correct. And yet, after 60 years, despite the truly massive advances in both hardware and software represented by projects like stable diffusion Minsky's problem remains far from solved. In practice, you can pretty much graph straight line between the simpler a task seems/earlier a it appears in the evolutionary enviroment, to how hard it will be to replicate. Playing chess is easy, Bipedal locomotion is difficult. Bipedal locomotion only seems easy to creatures like you and me because we've been doing it since we were two-years-old, and our ancestors spent millions of years refining the techniques and bio-mechanics that were bequeathed to us as infants.

What does this have to do with anything? My answer is that I feel like a recognition/understanding of Moravec's Paradox is one of the major components of inferential distance between myself and most others both in the rationalist movement, and in academia. It is why I am reflexively skeptical of grand unified social/political theories. and It is also why I remain deeply skeptical of GPT and the oncoming AI apocalypse it allegedly represents.

One claim you'll see guys like Elizer Yudkowsky, Bryan Caplan, and posters here on TheMotte make on a semi-regular basis is that "GPT knows how to play Chess". But if you press them on the topic, or actually look at chess games that GPT has played it becomes readily apparent that GPT makes a lot of stupid and occasionally outright illegal moves (eg moving rooks diagonally, attacking it's own pieces, etc...). What this demonstrates is that GPT does not "know how to play chess" at all. At least not in the same sense that Deep Blue or my 9-year-old can be described as "knowing how to play chess", or AlphaGo can be described as "knowing how to play Go".

Furthermore, once you start digging into their inner workings this lack of "knowing" appears to be a fundamental weakness of the Large Language Model architecture. At the end of the day it's still just a regression calculating the next most plausible word (or in the case of GPT-4 string of words) based on the correlations found in it's training data. Granted GPT-4 is certainly a step up from GPT-3 in terms being able to pass as human. The shift towards correlating longer statements rather than individual words seems to have plastered over a lot of the jarring discontinuities that made GPT-3 generated posts so easy to pick out. In contrast GPT-4 can actually kind of pass for human from the proverbial 50 ft away. Unlike prior GPT iterations, identifying it actually requires a level of careful reading or some sort of interaction.

Eugene Volokh's posts on Large Libel Models probably deserves a discussion of their own but INAL and not really interested in questions of liability. In any case he ends up running into the same issue with GPT that I did. Users here talk about instances of GPT "lying" or "hallucinating" and how to reduce the frequency of such instances, but the conversations inevitably devolve into self-referential nonsense because neither of these terms really describe what is actually happening. In order to "hallucinate" one must first be able to perceive. In order to "lie" one must first understand the difference between true and false. and GPT possesses neither. Simple fact is ask GPT for five examples of prosecutorial misconduct complete with citations and newspaper quotes and it will provide the names of five prosecutors, their alleged crimes, some juicy quotes, and supposed case numbers. However while the names provided might actually be real prosecutors, and the media outlet quoted might be a real outlet, if you actually look up the court records or try to find the quotes you're going to come up short because the example was not something that was pulled out of memory and provided, it was "generated" form the prompt in exactly the manner that a Large Language Model is designed to do.

to be continued...

edit: fixed link

But if you press them on the topic, or actually look at chess games that GPT has played it becomes readily apparent that GPT makes a lot of stupid and occasionally outright illegal moves (eg moving rooks diagonally, attacking it's own pieces, etc...). What this demonstrates is that GPT does not "know how to play chess" at all.

Imagine a blind person, without any sense of touch or proprioception, who has only heard people talk about playing chess. They have never seen a chessboard, never picked up a rook, and the mere concept of moving pieces is completely foreign to their sensorium.

And yet, when pressed, said person is able to play mostly legal moves, all the while holding the entire state of the board in their head. Correspondence chess via Chinese telephone.

I think anyone who witnessed such a feat would be justified in being highly impressed, whereas you happen to be the equivalent of someone complaining that a talking dog isn't a big deal because it has an atrocious accent, whereas Yudkowsky et al are rightly pointing out that you can't find a better way of critiquing a talking dog! Especially a talking dog that gets ever more fluent with additional coaching, to the point that it knows more medicine than I do, understands quite complicated math, and in general does a better job of being a smart human than the average human does.

In a park people come across a man playing chess against a dog. They are astonished and say: "What a clever dog!" But the man protests: "No, no, he isn't that clever. I'm leading by three games to one!"

Do the dogs not speak wherever it is you are from?

Part of my point is that computer programs being able to play chess at or above a Human level has been the norm for close to 40 years now. I would argue that the apparent inability to match that capability is a step backwards

There's around 0 dollars to be made by making a chess bot better than Stockfish. The days of rolling them out to spank the human pros is long gone, they just get up and start running for the hills when you pull out even the kind of bot that runs on an old smartphone.

In contrast, an AI that does tasks ~80% as good as a professional can, for pretty much all tasks that involve text, is economic disruption in a tin can. (Emphasis on professionals, because it is genuinely better than the average human at most things, because the average human is an utter humpty)

Notice how I said that it's a better doctor than me? Consider how much we spend on healthcare, one of the thousands of industries about to be utterly disrupted.

In contrast, an AI that does tasks ~80% as good as a professional can, for pretty much all tasks that involve text, is economic disruption in a tin can

But the difference is still in the tails. The top 1% is where the money is made in any competitive industry. That is why top tech companies are so obsessed with talent and recruiting. That is harder to automate than the rest.

Notice how I said that it's a better doctor than me? Consider how much we spend on healthcare, one of the thousands of industries about to be utterly disrupted.

It can automate the diagnosis process based on some input of symptoms, but other parts harder, like treating. Same for invasive tests and biopsies. Ai will disrupt it in some ways, but I don't think it will lower costs much.

I think you're adopting too much from a programming background when it comes to productivity. 10x programmers are far more common than 10x doctors or lawyers, because it isn't nearly as feasibly to simply automate the gruntwork without hiring more junior docs/lawyers.

I would say that productivity in the vast majority of professions is more along the lines of the Pareto Principle, such that a 80% competent agent can capture a substantial chunk of profits.

And what exactly is so hard about treatment? An AI doctor can write drug charts and have a human nurse dispense them. Invasive tests and biopsies are still further away, but I full believe that the workload of a modal doctor in say, Internal Medicine, can be fully automated today without any drawbacks. The primary bulwark against the tide is simply regulatory inertia and reflexive fear of such unproven advances.

Is there a good AI substitute for clinical examinations at present, or are we going to rely on patients self-examining?

I can honestly buy that in the short-medium term AI would take a better history and get differentials and suggest treatment plans better than the modal doctor. I could even buy that within that timeframe you could train AI to do the inspection parts/things like asterixis, but I don’t know how you’d get an AI to…palpate. Movement and sensation etc. are quite difficult for computers, I am to understand.

Alternatively maybe they’d just get so fucking good at the rest of it that professional examinations aren’t needed anymore, or that some examination findings can be deduced through other visual/etc means…

You'd be rather surprised at how little doctors palpate, auscult etc in practise. They're most used for screening, if there's any notable abnormality they get sent off straight to imagining instead of simply relying on clinical signs as was once common. It certainly plays a role, but with robots with touch sensors, it's hardly impossible to have AI palpate, it's just a skill that's rapidly becoming outmoded.

Oh I know well how doctors don’t do the things they teach us to do in medical school! But it did seem like one thing that they can’t (that easily) but we can (relatively easily), due to it being more of a physical and tactile thing.

That said, I find that I do examine people at least a few times a day.

I agree it’s hardly impossible but I’d be surprised if it wasn’t markedly harder to train?

More comments

If GPT hallucinates a wrong and deadly treatment, who do you sue for malpractice?

Right now? Nobody, because it's not licensed for medical use and uses massive disclaimers.

In the future when regulators catch up and it's commercially deployed and advertised to that end? Whoever ends up with the liability, most likely the institution operating it.

I see this as a comparatively minor roadblock in the first place.