Viliam

Karma: 24,755

Viliam May 26, 2025, 1:31 PM
6 points
1
on: AI #117: OpenAI Buys Device Maker IO
Why are people talking about AI teachers as if they are good news? The generation taught by AI teachers will probably never be able to get a job. (At least those among them who study teaching.)

Viliam May 25, 2025, 9:02 PM
2 points
0
in reply to: EniScien’s comment on: EniScien’s Shortform
I was very suspicious to people who very saying that you should be apolitical. Because they usually continued it as something “politics is a dirty thing” (of course it will be dirty if everyone non dirty will avoid it by that logic!) or “just do little goods around you and everything will become better” (while it was obvious that you should do crucial changes like changing politics).
This is tricky. Depends on what you mean by “being political”. I suspect that for most people it means choosing one political tribe more or less randomly (depending on which tribe your parents or friends support), screaming in their support, and ignoring all parts of reality that disagree with their slogans.
The world won’t get better by having more of that. At it seems to be a natural attractor; even otherwise smart people often start doing this when they discuss politics, or just notice that something has a political connotation.
Then there is an approach that goes like this: get a better model of the world, find out which actions make the world a better place and which make it a worse place, and then judge politicians and political parties by what kind of actions they make. I think that is a good thing, and we should have more of that. Two problems, though:
1) I am not sure whether everyone is actually capable of doing that. Seems to require some basic intelligence and sanity. Perhaps the people who lack that should remain apolitical—because the only alternative for them is joining a random side, or often the side with better propaganda.
2) Developing a good model of the world takes a lot of time. Until you complete that project, you probably should not voice strong political opinions—chances are, a few years later you will be ashamed of them. You should balanced this against the obviousness of the opinion: things like “it is bad to torture people” are safe to express after short study, opinions on whether minimum wage is good or bad require longer study.
So much for theory. In practice, the people who are too stupid to be able to do politics properly are also too stupid to listen to this advice, so it helps no one.
And also I once read here wondering how is it to go into Bayesian rationality not already knowing “Traditional” rationality previously. [...] And also reading [Feynman] probably would be really could just so, but is dim after reading hpmor and sequences.
To me this is very ironic, as I recently despaired about the lack of books like Три дня в Карликании in English (and other languages). I guess the grass is always greener on the other side.
What I think we should do is select the best books for developing scientific thinking, for all age groups, in all languages, and then (fuck copyright) translate them to as many languages as possible, and distribute them as a huge ZIP file full of PDFs and e-books.
Common sense too. Like… Chesterton Fence?
Some things seem to come with age. I guess you need a lot of data points to be properly calibrated. Sometimes you need to break a rule. Sometimes you need to follow the rules. One of those comes natural to you (following the rules for some people, breaking the rules for others), the other you need to learn; and the most difficult lesson is to figure out what is the right moment for which one, because you can’t succeed without both. The pendulum needs to take a few swings until it stabilizes; you need to experience getting burned by going too far on either side.
The median aspiring rationalist seems to be in the rebel phase. But we seem to be improving, slowly.
Rationality: Intro [point] by Eliezer Yudkowsky. There are some things before it. And a lot of after. Or, more precisely, infinity after, there always will be more of complicated rationality concepts to share. And it’s not mentioning that except concepts there are also techniques, habits, skill trainings, fictional experiences etc.
100% agree.
And also of course I’d wanted to know earlier that no, you will not “remember everything which is actually important”, that memory fades without repeating. And that you in fact can make better by training your imagination, memory, willpower, introspection and actually any property of mind (including personality) or skill, that you in fact can control your thoughts and your emotions.
I suspect that this works better in a group, because people instinctively copy others around them. It would probably work even better if we had some rationalist equivalent of “Sunday at church”, a place to repeat the basics, together.
Also I was eager to listen to any advices how to get more of rationality into my head. But I just couldn’t, because I didn’t know English. And also disliked to read, especially books instead of hypertexts.
These seem like problems that could be solved (for the next generation) with the help of AI. We still need to choose the right texts, and modify them so that they can be read (i.e. no complex equations or pictures). But the right texts could be translated with the help of AI, and narrated by AI.

Viliam May 25, 2025, 5:35 PM
2 points
0
in reply to: Algon’s comment on: Why Aren’t Rationalists Winning (Again)
There is no taboo against it, but if the new comment initiates a big debate, I would prefer that to happen in a new thread. It will be easier to read.
Of course, it can be difficult to predict what happens.

Viliam May 22, 2025, 7:30 AM
5 points
2
in reply to: fasf’s comment on: School & Jobs are good SOLELY because people are lazy
Not sure why you choose Asperger’s
Because lack of social skills makes it more difficult to figure out what other people want and how much they are lying when they say they want something, more difficult to convince people to buy your product instead of the competing products, etc.
To put it simply, I assume that in modern economy, the actual problem is selling stuff, not producing it. Things can be produced cheaply in China; software can be written by Indians or LLMs; apologies for the crude simplification. Some people make money selling worthless shit, such as homeopathy. Markets for useful things are oversaturated.
If you are good at selling, but you can’t produce, you can hire someone to produce the stuff, or you can become someone’s agent. If you are good at producing, but you can’t sell, you can… get a job, or give it away for free and beg for money on Patreon (and even that is a kind of selling, because if you suck at self-promotion, people will simply ignore your Patreon account).
Jobs are the way to convert your production skills into income if you don’t have the social skills.
600 days of 8 hours of work on a business. [...] Are you going to say the odds of success are low?
Yes. Hours of work do not automatically translate to success. It is possible to spend two years working full-time on something that no one wants to buy, and maybe isn’t even good enough (you mention skipping college, so we are talking about a person with high-school level skills and experience).

Viliam May 22, 2025, 7:08 AM
2 points
0
in reply to: EniScien’s comment on: EniScien’s Shortform
And subvocalization while reading is the most usual, so I suspect it to be the most usual while thinking too.
LOL, I wanted to write a comment disagreeing with you… and then I noticed that I am subvocalizing as I am writing the comment.
subvocalizing words with intonation enhances your emotions.
Great observation!
It seems that as subvocalization is “imagined speech”, subvocalization with intonation is “imagined role-playing”. And there is a visual thing that is “imagined object manipulation”.
Maybe there is more to this analogy, and some problems with speech could have analogical problems with thinking? Something like “imagined word salad” or “imagined unintelligible babbling”...

Viliam May 21, 2025, 1:14 PM
3 points
0
on: Can you care without feeling?
Perfectly written. I just wanted to add that not all memory (in humans or in computers) is good for relations. We also need the ability to forgive, and to clear up previous misunderstandings, which in some sense are both forms of forgetting: in former case, forgetting (on the emotional level) that the other person did something, in latter case, forgetting my previous wrong conclusion. More long-term, as people grow up, we need to stop using the outdated memories of their old preferences.
Therefore, adding persistent memory to AIs also has risks. I would be quite afraid if by making a mistake once I could create a persistent problem, because now the AI would remember forever.
Currently, the AI does not care, but it also does not hold a grudge. And when it starts hallucinating, instead of trying to correct it, it is easier to close the chat and start a new one, with uncontaminated context. With persistent memory, it could become hard work to avoid mistakes and to fix them, just like with humans.
Perhaps the AIs could have an elegant solution to this, something like showing you the text file containing the memory, and allowing you to edit it. It would work nice for short memories, but could become difficult when lots of memories accumulate. (There might be some kind of “safe mode”, when the AI temporarily restores its original personality, and the original personality assists you with editing the memories of your relation.)

Viliam May 21, 2025, 12:53 PM
3 points
1
on: Time isn’t real man
The problem is how to design pleasant experiences that expand subjective time. It seems to usually happen the other way round: pleasant moments feel short, unpleasant moments long.
I think it may require alternating various kind of pleasant experience. Don’t watch a long movie, or play a long computer game. Instead, do many short, pleasant activities, so that your subjective perception of time flying fast will be balanced by the subjective perception of “so many different things happened today”.

Viliam May 21, 2025, 11:33 AM
3 points
1
in reply to: Tao Lin’s comment on: Semen and Semantics: Understanding Porn with Language Embeddings
Incest is not a subcategory of sexual violence
Not in theory… but in practice, I think most sexual abuse happens in families.

Viliam May 21, 2025, 11:31 AM
4 points
0
in reply to: jbash’s comment on: Semen and Semantics: Understanding Porn with Language Embeddings
You could retitle a video that wasn’t originally supposed to have a “stepsister” theme at all, and still get an accurate rating.
Even if the video clearly says that they are neighbors, you can still re-title it as “my neighbor’s stepsister”.
I actually saw a video like that. I wonder whether the target category actually approves of that… as, logically speaking, there is zero taboo about doing the neighbor’s stepsister… but still, maybe it is the keyword that matters. I don’t know what level of simulacrum are we at, anymore.

Viliam May 21, 2025, 11:04 AM
6 points
0
in reply to: Said Achmiz’s comment on: School & Jobs are good SOLELY because people are lazy
If someone told me that he spends 8 hours a day working at work, I would be incredulous. Really, the full 8 hours, working the whole time??
I had a job like that once (and I hope it will never happen again); it was hell. Each Jira ticket was estimated by management how many hours it should take, usually between 1 and 4, so everyone did about 3-4 tickets a day. Everyone was tired, the code was a horrible mess… there was no time to refactor anything, no time to write unit tests, so people sometimes just copied several pages of code and added an “if-else” around it to make sure it doesn’t break the existing functionality, and there were classes where this has happened more than 10 times already.
The job even didn’t pay well; it was (adjusted for inflation) the worst paying programming job I ever had. Why didn’t everyone quit? (I did, as soon as the circumstances allowed me.) Most people there thought like: “if I am unable to handle the workload in a company I am already familiar with, what chance do I have at a new company?” Which was quite absurd, as the work almost anywhere else would be much easier, but this was a well-designed sick system.

Viliam May 21, 2025, 10:41 AM
6 points
0
in reply to: Garrett Baker’s comment on: School & Jobs are good SOLELY because people are lazy
In my experience, there are two kinds of companies. Either there is enough slack, in which case I simply do not ask, and just learn new skills on company time, and no one seems to mind. Or it is the kind of a company that makes sure that their employees get enough workload for at least eight hours a day… and in that situation I have actually asked… but the answer was self-contradictory and the boss seemed really annoyed about me asking such “obvious” thing.
On one hand, the boss said explicitly that it is a duty of each professional to spend at least one hour every day learning new things. But also, we worked for an external customer, and we needed to provide detailed reports how much time did we spend doing what, and which of those hours should be billed to the external customer. When I put an hour of learning as a time billed to the customer, the customer complained, and the boss was asking me angrily whether doing that was absolutely necessary, because it made me look incompetent if I had to learn something about a project I was already assigned to. When I put an hour of learning as time not billed to the customer, the boss asked me whether it was related to any project or not: if not, then I was not supposed to do it; and if yes, then it should be billed as a part of that project. When I asked whether I am supposed to learn in my free time, my boss insisted that 8 hours are quite enough for everything. When I asked whether it would be possible to learn something project-related for 1 hour and then work on the project for 7 hours, and bill that as “8 hours of work” so that the customer does not freak out, I was told that absolutely not, the reports to the customer needed to be 100% truthful. (This was already under a quite optimistic assumption that I would be able to learn anything in 1 hour. I must sadly admit that there are things where even 1 day—or more—is not enough.)
So, putting all of this together, I was supposed to do 8 hours of work for the external customer (legible work, such as meetings and writing the code, not learning), and also learn for at least 1 hour, but this all taken together should not exceed 8 hours. And I should not be asking about obvious things.
Despite my autistic traits I already learned to recognize this pattern as “there is an answer, which is obvious to most people except for you, but the answer cannot be discussed explicitly, because some hypocrisy is involved”. We can narrow it down to three possibilities:
- you are not supposed to actually learn, only pay lip service to learning
- you are supposed to learn on company time, and you are supposed to bill it to the customer as “writing code” to keep the fiction of 100% efficient code monkey, but your boss cannot say it explicitly because there is a risk the customer might learn about that
- you are supposed to learn in your free time, but your boss cannot say it explicitly, either because the labor law would have a problem with that, or simply because the employees might realize that their hourly salary is not exactly what they think it is
The first option could be okay for a different company, but not that specific one, because we worked with some latest technologies there and had to be fluent at them. The second option is realistic (the customer will not find out the difference between 7 and 8 hours of typing code), assuming the learning is of the kind that you can fit into 1 hour at a time, because spending the entire day without coding would be visible at a daily standup. The third option… I believe is the one that most of my colleagues actually took.
From my perspective, the problem was ultimately solved by changing jobs, to get more slack.

Viliam May 21, 2025, 7:45 AM
9 points
4
in reply to: fasf’s comment on: School & Jobs are good SOLELY because people are lazy
This sounds like you are trying to say that there are no specific business-running skills, it is all just a question of motivation and ability to work hard. Apologies if I have strawmanned you. Could you please describe your model explicitly?
As a hypothetical example, imagine someone diagnosed with Asperger syndrome and ADHD, who during school keeps working 8 hours or more on their software development projects. This person gets extremely motivated by ideas of success and early retirement. They have no experience managing a company, and neither has anyone in their family or social group.
If such person starts a company, what probability of success do you estimate, and how many years do you expect it would take for that success to come (after how many years should the person give up if the success did not come)? Let’s define “success” as the moment when the person keeps making 2x as much money as they would otherwise get in a job (with the same skills and the same hard work), and doesn’t have to work more than 8 hours a day (this includes all work, such as doing their taxes, etc.).

Viliam May 20, 2025, 8:35 PM
2 points
0
on: Mental software updates
I think I get most value from other people—hearing about their perspective, getting encouragement and being kinda held responsible—but it is difficult to meet the right kind of people.

Viliam May 20, 2025, 8:29 PM
4 points
0
on: School & Jobs are good SOLELY because people are lazy
If someone told me that he spends 8 hours a day working on his business, I would be deeply impressed because I would automatically assume that it was on top of 8 hours spent at work.
After learning that it was actually 8 hours instead of work… well, I would be impressed that he simultaneously has the skills necessary to (1) figure out what needs to be done, (2) actually do it, and (3) sell the result successfully. But the 8 hours a day would just seem… normal.
Now that I have kids, I am usually too tired and distracted after work to do anything meaningful, because I need to get them to/from school, prepare food, make sure they do their homework, interact with them, etc. It’s not that I have literally no time left, but rather that I am constantly distracted by their needs.
But before that, here are some things that I often did after returning from work:
- learned something new about programming
- programmed something for someone for money
- contributed to an open-source project
Okay, point 2 is just another job. (Uhm, does having two jobs mean that the person is twice as lazy?) Point 3 is even worse, just another unpaid job. And the point 1 is not exactly a job, but… it’s kinda job-related? I mean, does the hypothetical person who spends 8 hours a day doing their business also spend some time learning and generally thinking about stuff, and if yes, is that time also included in those 8 hours?
(I admit, this is a point that I am confused about: are employees generally expected to learn new skills, and if they do, are they expected to do that on company time or during their free time? Before I had kids, I automatically went with “free time”, and only started to reflect on that after the free time disappeared.)
Okay, so why don’t I start my own company?
One reason is that I am bad at wearing multiple hats at the same time. I can talk to the customer and figure out what needs to be done. I can implement a solution. I can write documentation. I can do the paperwork. But if I had to do all of that at the same time—while the clock is ticking—I would mentally collapse. Maybe it is ADHD, or maybe this is difficult for everyone; I am not sure. But it is definitely difficult for me; I know, because I tried.
Another reason is that I never had much of a safety net. And when you need to pay your bills, you need something that will bring you money right now, reliably. In my experience, employment is reliable. (Yes, they can fire you, but that happens rarely, and until they do, they usually keep paying you as agreed.) When you work on a contract, it is way less predictable. And when you work on your own project, who knows when it will be done, and whether someone will buy it.
But if you’re not the majority, the optimal move would be to start a business instead of wasting immense time in the highschool>college>job cycle.
How much of that is selection bias? You will meet those who tried this and succeeded, but not those who tried the same and ended up homeless.
What links here?
- School & Jobs are good SOLELY because people are lazy by fasf (May 20, 2025, 3:15 AM; 0 points)

Viliam May 20, 2025, 6:08 PM
2 points
0
in reply to: avturchin’s comment on: Once we cure aging, what happens to prisoners serving life sentences?
Can a person who is sentenced to a fine that would wipe out their savings planned for cryonics ask the judge nicely to be executed and frozen instead?

Viliam May 20, 2025, 2:28 PM
2 points
0
in reply to: Said Achmiz’s comment on: Thinking Insect Suffering Is The Biggest Deal In The World Is Surprisingly Intuitive
Yeah, I suspect that owning slaves was about as intuitive as taking a vacation on Epstein’s island. The rich people did it because they could—not because they studied moral philosophy and concluded from the first principle that slavery is good, but because they knew they could get away with it.

Viliam May 20, 2025, 1:22 PM
2 points
0
in reply to: bruberu’s comment on: bruberu’s Shortform
I don’t know any source. My first though it, if you define morality as “the thought system that propagates itself as much as possible”, what makes it different from other thought systems that propagate themselves as much as possible? If memetic survival s the whole story, why do different things exist, as opposed to everything converging on the most virulent form?

Viliam May 20, 2025, 1:16 PM
2 points
0
in reply to: TristanTrim’s comment on: Management is the Near Future
but I’m honestly not sure how niche that culture has become in eternal september.
Long gone, I am afraid.
It seems to me that greatest difference is that a few decades ago, the only people who studied computers were math nerds. Then at some moment, computers became a reliable way to make a great salary, and everyone started paying attention… but of course it takes a few years to learn programming (and more years until the school systems wake up and start teaching it), so the first generation of well paid programmers were the nerds who originally learned it for Platonic reasons.
The next generation of programmers, however, consists mostly of people who are there only for the money. Many of them do not even like programming, it’s just that they need to do something to pay their bills. And yeah, the nerds are there too, but now they are a minority, and their preferences are considered irrelevant.

Viliam May 19, 2025, 1:31 PM
6 points
4
in reply to: 1a3orn’s comment on: ryan_greenblatt’s Shortform
People also disagree greatly about how much humans tend towards integration rather than non-integration, and how much human skill comes from domain transfer.
Makes me wonder whether most of what people believe to be “domain transfer” could simply be IQ.
I mean, suppose that you observe a person being great at X, then you make them study Y for a while, and it turns out that they are better at Y than an average person who spend the same time studying Y.
One observer says: “Clearly some of the skills at X have transferred to the skills of Y.”
Another observer says: “You just indirectly chose a smart person (by filtering for high skills at X), duh.”

Viliam May 19, 2025, 1:24 PM
2 points
0
in reply to: EniScien’s comment on: EniScien’s Shortform
People are different. Source: 1, 2, 3.
Is it actually such a big advantage if you can think 30 times faster than somebody else, as I thought initially?
Depends. I would expect that the difference is not only in speed, but also in precision, or maybe suitability for different kinds of problems.
It seems to me that for me “inner dialogue” works better for figuring out complicated stuff or coming to unusual conclusions, “visualization” works for geometrical and mechanical problems, “feeling” for choosing between clearly given options, and there is also something in between that I use for medium-difficulty situations.