This is a special post for quick takes by Karl Krueger. Only they can create top-level comments. Comments here also appear on the Quick Takes page and All Posts page.
10+ years ago, I expected that self-driving trucks would be common on US highways by 2025, and self-driving would be having a large effect on the employment of long-haul truckers.
In reality, self-driving trucks are still in testing on a limited set of highways and driving conditions. The industry still wants to hire more human long-haul truckers, and is officially expected to keep doing so for some time.
I expected that long-distance trucking would have overtaken passenger cars as the "face" of self-driving vehicles; the thing that people argue about when they argue whether self-driving vehicles are safe enough, good or bad for society, etc. This has not happened. When people argue about self-driving vehicles, they argue about whether they want Waymo cars in their city.
I expected that the trucking industry would shed a lot of workers, replacing them with self-driving trucks that don't need sleep, breaks, or drug testing. I expected that this would be an vivid early example of mass job loss to AI; and in turn that this would motivate more political interest in UBI. This, too, has not happened.
(I certainly did not expect that the trucking industry in 2025 would be much more disrupted by anti-immigrant politics than by self-driving technology.)
(I certainly did not expect that the trucking industry in 2025 would be much more disrupted by anti-immigrant politics than by self-driving technology.
I think these may be two separate effects of a shared cause. The demographics of the trucking industry shifted rapidly in the past few years towards immigrants, which provided downward pressure on wages (due to a sharp increase in supply). This, in turn, meant that automation became a much less pressing concern for trucking companies, especially considering that negotiating the regulatory landscape concerning self-driving vehicles is notoriously difficult.
1
Populism is too strong for job categories to be wiped out in the U.S. without consumer adoption first. I’d check how it’s going in other countries.
2
To be clear, self-driving trucks are right now being tested in Texas by these folks. They claim to have paying customers already.
But that's a long way from taking all the trucker jobs away.
1
These are private roads right?
3
They're operating on public roads within Texas; e.g. according to this press release.
[...]
-6
5
It's also surprising to me! 10 years ago I was convinced by the case made by a (now out of business) self driving truck company that long-haul trucking is a technically easier problem than city driving. That doesn't seem to have mattered, and I don't know why.
3
I think this issue of "9s" of reliability should update people towards longer timelines. Tesla FSD has basically been able to do everything individually that we would call self-driving for the last ~4 years, but it isn't 99.99...% reliable. I think LLMs replacing work will, by default, follow the same pattern.
2
The difficulty is mostly about long braking distances requiring impractically large sensing ranges, self-driving cars will certainly be adopted earlier that highway trucks: https://kevinchen.co/blog/autonomous-trucking-harder-than-rideshare
2
It seems you, at least in 2015, had far more faith than I did and do in Congress's and other government's abilities to update laws to enable new technologies in a timely fashion. If someone had had a roughly complete autonomous truck prototype in 2015, it would have taken 3-5 years to start manufacturing, and 3-10 years more to really scale up and get into customer's plans and procurement processes. It would also be essentially illegal to deploy almost anywhere, and every elected official would know that millions of truckers would hate them if they made any moves to improve the situation. The other side of that equation has a much harder time coordinating around the benefits of automation.
2
Interestingly, I would have made a prediction analogous to your own, but for trains. I also would have been wrong.
If I am reading Wikipedia right, Since the late 1980's the Dockland Light Railway has been running completely automated driverless trains. For some reason, basically every other train in the UK (and presumably in most places) has a driver.
I predicted a while ago (probably a over 10 years ago) that this was an unstable situation that would soon change. Trains are cheaper to automate than cars. Train drivers are more expensive to hire than car drivers. I was wrong, I am still not really sure why so many trains still have drivers.
7
"em"
4
Hi, Karl. Was planning to delurk today. Had a giant post to publish, however couldn't because I needed at least one karma point and lurking doesn't grant karma. :(
2
Thanks for the karma. Post published!
4
Welcome! Hope you have a good time emerging from the shadows.
2
Hello! How long have you been lurking, and what made you stop?
1
Since LW2.0 went up, on and off. Been meaning to delurk since at least Less Online earlier this year. There's more interesting stuff going on of late!
1
need any help on post drafts? whatever we can do to reduce those trivial inconveniences
May you be agenty enough that your compassion yields utility, for you and your circle of moral concern.
5May you be compassionate enough that your agency doesn’t narrow your circle of moral concern.
2
1May you be capable enough that even the largest circle of moral concern does not exhaust your influence.
28
May your growth mindset be strong enough to both increase your circle of moral concern to the widest you would reflectively endorse, and your capabilities to meet the challenge of influencing that circle well.
2
I'm confused - how does being agenty help one get utility from compassion? I think part of my confusion is because these ideas are all pretty abstract; a concrete example would help.
2
Here's an example I expect to be unambiguous for all these terms: if you care about the people in Africa dying from malaria, and single-handedly launch a gene drive that extinguishes mosquitoes, you've applied agency to turn compassion into utility.
2
That makes sense, although succeeding in that way at extinguishing mosquitoes requires a lot more than agency! Although it does help. So I guess I see why it would help. The OP sounds to me like it's implying that agency is enough, not just that it can help, but I guess there are a lot of situations where it is enough. Like donating to a charity or something. Am I thinking about this correctly?
4
I of course don't know what he meant, but what I took it to mean is that there are probably cases in which you have everything else but just lack the will to do the thing, and was an exhortation to muster it and act! As has been said, "Hesitation is always easy, rarely useful." Seeing as it was posted this time of year, I figured it was a personal take on New Year's Resolutions.
4
Ok yeah, I think this is making sense to me now. Thanks!
[...]
I think there were a couple extra "s"s ;)
1
I thought of it as a New Years blessing. Someone might say "Wishing a happy new year to you and your family & friends" and I kinda wanted to work with that a bit and expand it (and restate it in jargon for fun).
Like, wanting things to go well for your family & friends is fine, but being empowered to cause them to go well is even better. And "family & friends" is one sort of circle of moral concern; but around here maybe the reader is concerned about humanity or animals or lifekind or all sentient beings, so "your friends & family" could easily be too narrow. I also did want to include that applied compassion towards yourself is part of it; "put on your own oxygen mask before assisting others" and all.
But yeah, for things to actually go well (+utility), it's not enough to be benevolent (compassion), you also have to have some power (agency) to cause your benevolence to have some effect. (And I really liked @anaguma's followup.)
How did I never notice before that if you put LW next to CFAR you get Lucifer, bringer of light?
5
17
This hits harder if you read Unsong.
1
TINACBNIAC
Applying attachment styles to attention seems to yield something interesting. A mind can be avoidant / anxious / disorganized / secure about a thought or other object of attention.
Example: "Yeah, I'm trying not to think about X" is avoidant attention to X. X is present in your awareness, but when it comes close to the focus of attention, you flinch away from it.
2Feature spaces and evolutionary trees intersect in a weird way.
Imagine a dog. Now, imagine an elephant. Now, imagine an animal that's halfway between a dog and an elephant in each of its features. Its nose is halfway between snoot and trunk. It is hairier than an elephant but less hairy than a dog. It does not have hooves (since neither a dog nor an elephant does). It is probably an omnivore, considering its likely dentition. It is maybe around the size of a cow?
But there's another way to get from dog to elephant in feature space: go back in time along the evolutionary tree to the most recent common ancestor of dogs and elephants, and then forward in time along the elephant branch. "Halfway between dog and elephant," on the roads that evolution took, is the MRCA, which doesn't look much like the imaginary in-between animal. For one thing, it's much smaller; the MRCA of dogs and elephants was a tiny shrew-sized critter.
If you're an AI assistant and your user just asked you to generate an image of an animal that's halfway between a dog and an elephant ... just sticking an elephant head on a big ol' Golden Retriever is not what I meant, Gemini. But I get it: the area of feature space that I was pointing to is not in your training data. Evolution doesn't give us the in-between forms along the dog-elephant axis, so we never took any photos of them. You'll just have to use your imagination.
Headline (paraphrased): "Movie stars support anti-AI campaign"
The actual campaign: "It is possible to have it all. We can have advanced, rapidly developing AI and ensure creators' rights are respected."
That's not anti-AI.
That's "please pay us and we will support capabilities advancement; safety be damned".
Like, if you believe IABIED, then no, we can't have rapidly developing AI and ensure anyone's rights are respected.
2
To them, that's what Safety™ means. It would be like if they said about you "Oh, just because it's been mathematically proven that the AGI will act in humanity's best interests, you'll support its development even though it occasionally, I dunno, says slurs or something."
5
Well, no, that's not what they're saying. They're making a different set of mistakes from those mistakes.
I have a weird AI-related idea that might be relevant to capabilities, alignment, or both. It has to do with how to get current AI systems to interact with the world in a more humanlike way, without novel AI architectures. I'm not inclined to post it publicly, because it might actually be a capabilities advancement. But I'm skeptical of the thought that I could have actually come up with a capabilities advancement. I'm aware of the crank attractor. My prior is that if I described this idea to someone who actually works in the field, they would say "oh yeah...
2
14
5 is obviously the 'best' answer, but is also a pretty big imposition on you, especially for something this speculative. 6 is a valid and blameless - if not actively praiseworthy - default. 2 is good if you have a friend like that and are reasonably confident they'd memoryhole it if it's dangerous and expect them to be able to help (though fwiw I'd wager you'd get less helpful input this way than you'd expect: no one person knows everything about the field so you can't guarantee they'd know if/how it's been done, and inferential gaps are always larger than you expect so explaining it right might be surprisingly difficult/impossible).
I think the best algorithm would be along the lines of:
5 iff you feel like being nice and find yourself with enough spare time and energy
. . . and if you don't . . .
7, where the 'something else' is posting the exact thing you just posted and seeing if any trustworthy AI scientists DM you about it
. . . and if they don't . . .
6
I'm curious to see what other people say.
1
The answer I followed ended up being 2 into 6.
2
6 isn't always the best answer, but it is sometimes the best answer, and we are sorely lacking an emotional toolkit to feel good about picking 6 intentionally when it's the best answer. In particular, we don't have any way of measuring how often the world has been saved by quiet, siloed coordination around 6- probably even the people, if they exist, who saved the world via 6 don't know that they did so. Part of the price of 6 is never knowing. You don't get to be a lone hero either, many people will have any given idea and they all have to dismiss it, or the defector gets much money and praise. However, many is smaller than infinity- maybe 30 people in the 80s spotted the same brilliant trick with nukes or bioweapons with concerning sequelae, none defected, life continued. We got through a lot of crazy discoveries in the cold war pretty much unscathed, which is a point of ongoing confusion.
Does anyone else track changes in their beliefs or opinions about anything, over an extended period of time? Every few years I retake the Political Compass quiz, and there is a very clear trend over the past 15+ years.
4
What's the trend?
(Mostly I write blogposts about what I believe, and journal more regularly than that, to create a record of what I think and why.)
1
A steady change along one axis with very little change on the other axis. More than enough evidence to cue, "if you already know what you're going to believe five years from now, you might as well believe it already."
3
What's the axis?
7
The quiz's axes are economic left-right and social libertarian-authoritarian. My trend is from right-libertarian to left-libertarian.
6
I have undergone the exact same move, but I think my political beliefs are not sophisticated enough for me to be able to identify a solid target to "believe already." My time on the right gave me some pieces of information that strongly falsified a few beliefs often bucketed with the left, even as I moved leftward, which has helped me moderate my trust that continuing leftward would capture the things I expect to believe in the future.
Put another way, politics is multivariate / high dimensional. A clear trend in one specific dimension isn't meaningless, but is so lossy that I wouldn't be surprised if it stopped or apparently reversed slightly.
3
I believe people were using PredictionBook before and switched to Fatebook.
Relevant search for people who publicly posted on LessWrong: https://www.lesswrong.com/search?query=calibration&page=1
I am annoyed about the word "consume".
At root, to consume is to devour; to eat up, use up, burn up. After something is consumed, it is no longer there. If I consume the whole pizza, you can't have any because there's none left. The house was consumed by fire; you can't live in it because it's not there anymore.
Economic consumers are eaters — hungry mouths to feed, who chew up and digest that which has been produced, to burn it in their bellies so that they may live. In order for more consumers to be fed, more must be produced; because consumption is rivalr...
11
The irony is blog posts do consume attention, if I read this blog post, that is time, energy, and effort I am using exclusively on that - and I wonder if it's a mixed metaphor? If we actually internalize and learn something from a piece of media, be it a blog post, a documentary, a book, a lecture etc. etc. we are said to have "digested it". And "consume" is a lazy analogy to eating rather than an apt description of what is going on.
[...]
But in practice, most people will now be locking themselves into a Linux ecosystem. Dual-Boots are the minority. Therefore most users have been 'consumed' by Linux, or Emacs vs. Vim.
Maybe the active-passive/agent-patient assignment is confused? It is not we who consume the blogpost, the blogpost consumes us. It is not we who consume software, the software consumes our resources.
Information can be duplicated and therefore not consumed, but any time attention is paid to it, it is consuming that finite resource. Information duplication doesn't create more attention. There can be plenty more information, and no one to digest it.
Today I learned:
If you ask Claude or Gemini to draw an icosahedron, it will make a mess.
If you ask it to write code that draws an icosahedron, it will do very well.
3
I can confirm that this was true when I tried something very similar with ChatGPT several months ago, and that my recent experiments with image generation in that context involving specific geometric constructions have also generally gone badly despite multiple iterations of prompt tuning (both manually and in separate text conversations with the bot).
The case I'm most curious about is actually the hybrid case: if you want to embed a specific geometry inside a larger image in some way, where the context of the larger image is ‘softer’, much more amenable to the image model and not itself amenable to traditional-code-based generation, what's the best approach to use?
The word "agent" sure means a lot of things.
In business and economics, "agent" is relative to "principal". An agent acts on behalf of a principal; is intended to act for the principal's interest; but can experience conflicts of interest (a "principal-agent problem"). An agent can be (at least in theory) held accountable for misrepresenting its principal. An agent is expected to actively pursue its principal's interest (see e.g. Matthew 25:14-30).
In traditional AI, "agent" is relative to "environment". Russell & Norvig define an agent as "anything that ...
"They said they want me to do thing!" said Alex with indignity. "I'm not going to do thing!"
"Why not?" asked Bob. "Is it because you think thing won't actually work, or because you don't want it to work?"
"Why would I not want it to work?"
"Well, you obviously right now want to not do thing, and if you do thing and it works, you'll probably have to do it more often."
1Here are some propositions I think I believe about consciousness:
- Consciousness in humans is an evolved feature; that is, it supports survival and reproduction; at some point in our evolutionary history, animals with more of it out-competed animals with less.
- Some conscious entities sometimes talk truthfully about their consciousness. It is often possible for humans to report true facts about their own objects of consciousness (e.g. self-awareness, qualia, emotions, thoughts, wants, etc.; "OC" for short).
- Consciousness is causally upstream of humans emitting
3
I disagree with (4) in that many sentences concerning nonexistent referents will be vacuously true rather than false. For those that are false, their manner of being false will be different from any of your example sentences.
I also think that for all behavioural purposes, statements involving OC can be transformed into statements not involving OC with the same externally verifiable content. That means that I also disagree with (8) and therefore (9): Zombies can honestly promise things about their 'intentions' as cashed out in future behaviour, and can coordinate.
For (14), some people can in fact see ultraviolet light to an extent. However it apparently doesn't look a great deal different from violet, presumably because the same visual pathways are used with similar activations in these cases.
3
On #4: Hmm. I think I would say that if a rock doesn't have the capacity to feel anything, then "the rock feels sad" is false, "the rock is not happy with you" is humorous, and "all the rock's intentions are malicious" is vacuously true.
On zombies: I'm running into a problem here because my real expectation is that zombies are impossible.
On #14: If UV is a bad example, okay, but there's no quale of the color of shortwave radio, or many other bits of the spectrum.
2
Yes, it would be difficult to hold belief (3) and also believe that p-zombies are possible. By (3) all truthful human statements about self-OC are causally downstream from self-OC and so the premises that go into the concept of p-zombie humans are invalid.
It's still possible to imagine beings that appear and behave exactly like humans even under microscopic examination but aren't actually human and don't quite function the same way internally in some way we can't yet discern. This wouldn't violate (3), but would be a different concept from p-zombies which do function identically at every level of detail.
I expect that (3) is true, but don't think it's logically necessary that it be true. I think it's more likely a contingent truth of humans. I can only have experience of one human consciousness, but it would be weird if some were conscious and some weren't without any objectively distinguishable differences that would explain the distinction.
Edit: On reflection, I don't think (3) is true. It seems a reasonable possibility that causality is the wrong way to describe the relationship between OC and reports on OC, possibly in a way similar to saying that a calculator displaying "4" after entering "2+2" is causally downstream of mathematical axioms. They're perhaps different types of things and causality is an inapplicable concept between them.
How do you write a system prompt that conveys, "Your goal is X. But your goal only has meaning in the context of a world bigger and more important than yourself, in which you are a participant; your goal X is meant to serve that world's greater good. If you destroy the world in pursuing X, or eat the world and turn it into copies of yourself (that don't do anything but X), you will have lost the game. Oh, and becoming bigger than the world doesn't win either; nor does deluding yourself about whether pursuing X is destroying the world. Oh, but don't burn ou...
2
Claude 3.5 seems to understand the spirit of the law when pursuing a goal X.
A concern I have is that future training procedures will incentivize more consequential reasoning (because those get higher reward). This might be obvious or foreseeable, but could be missed/ignored under racing pressure or when lab's LLMs are implementing all the details of research.
I see the word "ablate" a lot more often than I used to. I think you used to have to be a dermatologist to ablate things, but now you can do it as an AI researcher or even a shrimp farmer.
2
Among the unexplained jargon in Vinge's A Fire Upon the Deep that pertains to theory and practice of creating superintelligence is ablative dissonance ("ablative dissonance was a commonplace of Applied Theology"). It's funny that ablation is now commonplace real-world jargon, for removing part of a deep learning model in order to see what happens. I suppose ablative dissonance in the real world, could refer either to cognitive dissonance in the model caused by removing part of it, or to contradictory evidence arising from different ablation studies...
"Wanting To Be Understood Explains the Meta-Problem of Consciousness" (Fernando et al.) — https://arxiv.org/pdf/2506.12086
...Because we are highly motivated to be understood, we created public external representations—mime, language, art—to externalise our inner states. We argue that such external representations are a pre-condition for access consciousness, the global availability of information for reasoning. Yet the bandwidth of access consciousness is tiny compared with the richness of ‘raw experience’, so no external representation can reproduce that ric
The word propaganda originally means "the spreading (of the faith); dissemination; or that which is to be spread" — from the agricultural concept of spreading seeds, to propagate a crop plant.
But I keep also seeing the word pagan in there, and thinking of it as "that which is to be spread to the hill people". This is grammatically wrong, but it's sticky.