Viliam

Karma: 24,646

Viliam May 10, 2025, 6:22 PM
4 points
1
in reply to: Richard Horvath’s comment on: Interest In Conflict Is Instrumentally Convergent
Trying to summarize the method:
- list all know facts
- list all competing theories
- make an M×N table, highlight places where the fact contradicts the theory
- require an explanation for each such place in the current theory
- if a new theory is made, add a new column to the table, and evaluate all cells in the new column
I guess this mostly avoids the failure mode when someone uses an argument A to support their theory X, later under the weight of evidence B switches to a theory Y (because B was incompatible with X, but is compatible with Y), and you fail to notice that A is now incompatible with Y… because you vaguely remember that “we talked about A, and there was a good explanation for that”.
The admitted disadvantage is that it takes a lot of time.

Viliam May 10, 2025, 5:42 PM
3 points
1
in reply to: Screwtape’s comment on: Interest In Conflict Is Instrumentally Convergent
Most security experts a bank would reasonably hire are not bank robbers, you know?
Yes, it would be useful to know how exactly that happens.
I suspect that a part of the answer is how formal employment and long-term career changes the cost:benefit balance. Like, if you are not employed as a security expert, and rob a bank, you have an X% chance of getting Y money, and Z% chance of ending up in prison. If you get hired as a security expert, that increases the X, but probably even more increases the Z (you would be the obvious first suspect), and you probably get a nice salary so that somewhat reduces the temptation of X% chance at Y. So even if you hire people who are tempted to rob a bank, you kinda offer them a better deal on average?
Another part of the answer is distributing the responsibility, and letting the potential bad actors keep each other in check. You don’t have one person overseeing all security systems in the bank without any review. One guy places the cameras, another guy checks whether all locations are recorded. One guy knows a password to a sensitive system (preferably different people for different sensitive systems), another guy writes the code that logs all activities in the system. You pay auditors, external penetration testers, etc.
There is also reputation. If someone worked in several banks, and they those banks didn’t get robbed, maybe it is safe to hire that person. (Or they play a long con. Then again, many criminals probably don’t have patience for too long plans.) What about your first job? You probably get a role with less responsibility. And they probably check your background?
...also, sometimes the banks do get robbed; they probably do not always make it public news. So I guess there is no philosophically elegant solution to the problem, just a bunch of heuristics that together reduce the risk to the acceptable level (or rather, we get used to whatever is the final level).
So… yeah, it makes sense to learn the heuristics… and there will be obvious objections… and some of the heuristics will be expensive (in money and/or time).

Viliam May 10, 2025, 4:26 PM
4 points
2
in reply to: Steven Byrnes’s comment on: steve2152′s Shortform
Yes, I meant it that way.
LLMs quickly acquired the capacity to read what humans wrote and paraphrase it. It is not obvious to me (though that may speak more about my ignorance) that it will be similarly easy to acquire deep understanding of everything.
But maybe it will. I don’t know.

Viliam May 9, 2025, 9:50 PM
3 points
1
in reply to: Yaroslav Granowski’s comment on: Yaroslav Granowski’s Shortform
In some sense, a man with a pencil and paper is much better mathematician than a man without them. And we have not yet reached the limits of how helpful the helpful tools could be. For example, there are various forms of note-taking software, and people debate endlessly about their advantages and disadvantages, which suggests that all of them are far from perfect. (And that’s still about personal notes. A perfect tool for collaborative note-taking would require even more functions.)
Okay, sounds potentially interesting, go ahead! (I hope you don’t mean zettelkasten.)

Viliam May 9, 2025, 9:44 PM
2 points
2
in reply to: Cole Wyeth’s comment on: ete’s Shortform
Yes, but enough bugs caught can be enough to switch the equation from “this is not worth doing” to “worth doing, even if we need to check everything twice”.

Viliam May 9, 2025, 9:39 PM
4 points
0
in reply to: Steven Byrnes’s comment on: steve2152′s Shortform
The world is complicated (see: I, Pencil). You can be superhuman by only being excellent at a few fields, for example politics, persuasion, military, hacking. That still leaves you potentially vulnerable, even if your opponents are unlikely to succeed; or you could hurt yourself by your ignorance in some field. Or you can be superhuman in the sense of being able to make the pencil from scratch, only better at each step. That would probably take more time.

Viliam May 9, 2025, 9:34 PM
20 points
4
on: Interest In Conflict Is Instrumentally Convergent
This is a horrible situation, where excessive knowledge of some bad action X could be evidence of being:
- the kind of bad actor who does X or plans to do it, or
- a person who was falsely accused of doing X, or
- someone who spends a lot of effort on protecting themselves or others from X, or
- a nerd who happens to be obsessed with X.
Taking all of them them together, we get a group that has the best knowledge of X, and which is dangerous to approach, because too many of them are bad actors. (Also, there is a possible overlap between the groups.)
Even worse, if you decide to avoid approaching the group and just study X yourself… you become one of them.
However, having zero knowledge about X makes you an easy victim. So what are we supposed to do?
I guess the standard solution is something like “try to learn about X without coming into too much contact with the teachers (learn from books, radio, TV, internet, or visit a public lecture), and keep your knowledge about X mostly private (do things that reduce the chance of X happening to you, maybe warn your friends about the most obvious mistakes, but do not give lectures yourself)” or sometimes “find a legible excuse to learn about X (join the police force)”. Which, again, is something that the bad actor would be happy to do, too.
They say that the former criminals make the most efficient cops. I believe it also works the other way round.
I guess Jordan Peterson would say that you cannot become stronger without simultaneously becoming a potential monster. The difference between good and bad actors is how much the “potential” remains potential. Weak (or ignorant) people are less of a threat, but also less of a help.
It could help if you know the person for a long time, so you could see how the excessive knowledge of X manifests in their actual life. Difficult to do for an online community.

Viliam May 9, 2025, 9:04 PM
4 points
0
in reply to: Yaroslav Granowski’s comment on: Yaroslav Granowski’s Shortform
As far as I remember, the following kinds of human-based superintelligences were discussed here:
- superbabies—genetically engineered humans, getting the best traits, including intelligence
- cyborgs—humans connected to computers
- uploads—humans (or just brains) simulated in a computer, neuron by neuron, only much faster
Relevant tag: intelligence amplification
Is any of this close to your idea?

Viliam May 8, 2025, 7:28 PM
4 points
0
in reply to: Cole Wyeth’s comment on: ete’s Shortform
I guess the professions that benefit most would be limited by two factors:
- they work with virtual stuff, not material objects
- mistakes don’t matter, because they are easy to notice
Translators get a huge multiplier, because you can skim the automatic translation and notice when something feels off. Software engineers can use unit tests. Who else is in this group?

Viliam May 8, 2025, 7:22 PM
3 points
0
in reply to: Morpheus’s comment on: Morpheus’s Shortform
Ah, there is that weird relation between writing and publishing. On one hand, publishing gives you the incentive to write. On the other hand, publishing gives you the anxiety about writing.
I have an intuition that a good solution to this would be some kind of multi-layered publishing system. Where you publish the article first to your close friends, and then if you feel okay about it, you promote it to larger audiences. At each step you can update the article; generally because you want to impress the larger audience more, but also based on the feedback from the previous audience. Probably each audience would have a separate comment section. At the first step, there would be no anxiety, because only your close friends will see it; and the next step, you would be encouraged by the feedback from the previous step.
Something similar could be achieved by having two blogs, one only accessible for friends. Maybe as a mutual help, for a small group of bloggers who want to overcome their writing blocks.
Topics that go into my head about what to write about revolve mostly about how it is hard to write.
Yeah, that is a trap. Although you might write an article like that, without publishing it, as a therapy.
I decided no actually this nested list doesn’t work on Twitter, so I will put this in an lw shortform
By the way, some of my LW posts started as shortforms that grew too long, so I decided to publish them separately. This reduces the anxiety a lot, because as long as I am writing in the shortform editor, I don’t really try to make it perfect. And when I switch to article form, I already have a large part of it written.
I wonder, maybe I should try writing all my article ideas as shortforms. To start writing, and commit to publishing it today, either as a shortform or as an article, depending on how well the shortform writing will go.

Viliam May 8, 2025, 6:12 PM
1 point
0
in reply to: gwern’s comment on: Logan Riggs’s Shortform
I didn’t see a single empty shelf.
In that case, as a communism survivor, I agree that there definitely was no toilet paper shortage.
lots of mediocre substitutes of the sort that were used before ‘toilet paper’ as a specialized category became the norm—like newspapers.
Newspapers? What is this, 20th century? Try using a smartphone. :D
But yes, paper towel rolls. (I don’t remember whether such thing existed during communism.)

Viliam May 8, 2025, 4:43 PM
2 points
0
in reply to: Logan Riggs’s comment on: Logan Riggs’s Shortform
Makes me feel confused about Economics 101. If people can easily predict that the toiler paper will run out, why aren’t the prices increasing enough to prevent that.

Viliam May 8, 2025, 4:37 PM
5 points
0
in reply to: Mandatory Topic’s comment on: It’s ‘Well, actually...’ all the way down
Sadly, this assumes that all good arguments are legible.
I think we would both agree that there is a nonempty subset of chemicals that is dangerous for humans to eat, and that it would be very difficult to provide an exact definition of that set.

Viliam May 8, 2025, 4:31 PM
4 points
2
on: Unexpected Conscious Entities
The countries are an interesting example. Yes, they can model other countries as agents and negotiate with them. But they can also negotiate with individual people in other countries (e.g. recruit them as spies). So we potentially have two levels of consciousness here, with the ability to communicate across layers.
As an analogy, imagine being able to communicate with other people’s cells. For example, you can’t defeat your rival, but you can convince some of his cells to give him cancer.
I guess something similar already happens, except instead of on the level of cells, it is on the level of sub-agents. For example, by being attractive, you kinda subvert the other person’s part of brain to argue in your favor. You can subvert the other person even more strongly, by becoming their drug dealer.

Viliam May 8, 2025, 4:19 PM
4 points
0
on: Global Risks Weekly Roundup #18/2025: US tariff shortages, military policing, Gaza famine.
Russia is expanding bases and planning to mass large numbers of troops along its borders with Finland and the Baltic states, in an indication that it may be preparing for war in the Baltics. A Russian invasion of one or more Baltic states would challenge NATO and would be a test of whether Article 5 still has teeth.
From the second linked article:
The Kremlin is planning to establish a new army headquarters in the Russian city of Petrozavodsk, about 100 miles east of the Finnish border, which will oversee tens of thousands of troops over the next several years. Many of these troops will be deployed to the area after the Ukraine war ends.
I think the obvious solution is to make sure that after the Ukraine war ends, there will be no Russian troops to return home from there.
That said, I don’t expect European politicians to do the obvious thing.

Viliam May 8, 2025, 4:07 PM
4 points
0
on: Global Risks Weekly Roundup #18/2025: US tariff shortages, military policing, Gaza famine.
Lack of toilet paper is something I associate strongly with communism.
You say it could be a self-fulfilling prophecy. But I guess that was also true during communism, to a certain degree. Whenever the leaders said “there will be no shortage of toilet paper, be responsible comrades and don’t panic”, of course everyone ran to the nearest shop and bought as much toiler paper as possible. That was the rational behavior, individually. (Especially because, unlike food, toiler paper does not spoil. So there is almost no downside to buying more than you need at the moment.)
In capitalism, in theory, some entrepreneurs should predict the shortages, and make a profit on that prediction by increasing production, increasing imports, and increasing costs. So instead of a national shortage, there should be a temporary national increase in costs, and some local shortages.
We’ll see.

Viliam May 7, 2025, 10:01 PM
2 points
0
in reply to: CstineSublime’s comment on: CstineSublime’s Shortform
I’ve been repetitively told that “you need to put your stuff out there” is true—but I don’t believe it. These people are often peers, stating it as a fact
Can they give you specific examples of the clients they gained as a result of publishing a video on a social network? (Maybe there was something special about those videos, and it is not true for videos in general. But maybe you could copy that special thing.)
Likes and comments don’t translate into commissions and clients.
Agreed, they don’t. Maybe shares make it more likely for the video to reach a potential client.
I suspect that a good video needs to be “actionable”: it should give a specific example of a problem that you can solve, and it should explicitly encourage them to contact you if they need to have a problem like that solved.
Other types of videos are only useful if they make people click on your video feed and find the “actionable” ones.
But that’s just a guess; I never had a job like that.

Viliam May 6, 2025, 11:19 PM
2 points
0
in reply to: testingthewaters’s comment on: testingthewaters’s Shortform
Quietly at first, then openly as people get used to it. You always want to have just slightly more ads than your competitors, because having much more could make people switch.

Viliam May 6, 2025, 11:16 PM
2 points
0
in reply to: CstineSublime’s comment on: CstineSublime’s Shortform
Things that can influence the belief:
- peer pressure: when other people say that something is true, the beliefs becomes contagious
- getting quick feedback: if doing a part of it already gives you a partial reward, it is motivating
- getting the information from a trustworthy source: if their previous claims turned out to be true...
- getting other reward in parallel, e.g. being socially rewarded for working on the project

Viliam May 6, 2025, 11:08 PM
2 points
0
in reply to: Nina Panickssery’s comment on: Prioritizing Work
I understand the deathbed scenario as a warning against procrastination—the kind where you actually may work a lot, it’s just that you keep postponing the important things, with the vague idea that you will get to that “later”.
The deathbed scenario is giving you the perspective when there is no “later” and your life ended up just doing the usual stuff. If you do not endorse such outcome, then you need to figure out how to include some of those important things in your life now rather than unspecified “later”. Otherwise, there is a risk that the perspective may actually be a correct prediction of your life.