igiveup

And I consider that a great achievement, not being on the HN front page when I have nothing to say.

This stuff should have some warning tag: rando's musings about LLMs. Lately, whenever I think I found an interesting article it turns out to be this.

22ns might be about 100 processor instructions. Somehow I doubt that any programming language can parse 50 strings in 100 instructions, let alone with naive code.

"Blessing of dimensionality"?

Oh, the OP pretty much told me that they didn't like taking notes. I am trying to understand why, what is the mechanism? Is it like burning out? Can it happen to me and my notes when I start doing something wrong? Why is "just do less of it" not a solution? Why is "just expect less from it" not a solution? Am I an exception? If so, why?

"All you need to understand" is for insects.

I don't get this. I have a megabyte or two of plain-text notes, and going through them and maintaining them and extending them is fun (apologize to the person who doesn't like threes). There are notes for a novel, ideas for future personal projects (way too many to do in my remaining lifetime), attempts at capturing my understanding of great scientific and philosophical problems, weird things I invented in my dreams, various ideas which didn't fit anywhere else. Guess what, the novel will probably never get written, projects will never get done, I will not make a philosophical breakthrough. So what?

Some ideas on how am I supposed to start hating my notes:

* They grow to 100MB, then it starts to be a burden

* I switch from notepad.exe to a dedicated application which somehow exploits my hobby of writing notes

* I develop OCD or something else

None of this seems very relatable. At this point, I might be writing a new note with these ideas, updating it when I get more ideas or when the one, most plausible explanation jumps out at me. Then I would read it years later and have something to think about before bed and have a good feeling that I didn't lose something and I am not left with thougths about the last episode of a TV show. Or is that supposed to be a bad feeling?

The point is minimizing curvature, not maximizing length. Curved path makes the accelerated particles lose energy through radiation.

Impossible with just the basic inductive principle of tokamaks, yes. Some years back I learned that you can keep the current going with microwaves, not sure about recent progress. You can also approximate steady state by reversing polarity of the warp - ehm - magnetic field regularly.

Yes, stellarators are different.

While temperature may be in the stellar ballpark, pressure should be much lower. That is fine because we are not trying to do proton-proton fusion (that one is very slow even in a star) but a much easier deuterium-tritium fusion.

Ability to perceive your own thoughts. Access to your own debug logs.

I think this is distinct from "ability to perceive green, which doesn't exist". A neural network trained to distinguish green in the output of a spectroscope will perceive green without ever knowing there is something like a "neural network" or "thoughts".

Also, this probably can be detected from the outside via debugging. What cannot be detected may be the thing that distinguishes a hypothetical "philosophical zombie" from a truly conscious human, but I don't think anything like a philosophical zombie exists. Once it is physically identical to the human, it will also be thinking identically.

As a next step, you may observe humans around you and realize that the thoughts you perceive seem to be running inside the head of one of these humans (which you will call "me"). However, I don't think knowing what you look like from the outside is necessary for consciousness.

What is an optimum in a two-objective (time, space) problem?

Also, how exactly does time help to save space? Is this about some sophisticated hash functions? Handling collisions?

Is that wrong? I don't know the logic of the game, but a 20:10 battle can very well have different odds than a 2:1 one.

To me, it sounds more like you're saying the same thing as me: It is not a law of nature, it is text on paper, a human fiction. Yet, somehow you manage to disagree with me. How did that happen?

Human laws do not follow laws of nature, or laws of logic. Trying to rationalize why something is or isn't owning or stealing is a misunderstanding.

"It is not missing anywhere, so it isn't stealing" does not apply. What applies is more like "This paper says it is stealing, so it is." The paper says it because a human wrote it there, as a result of whatever complexity going on in their natural neural network.

Please notice I didn't say "right", "wrong", "good", or "bad" anywhere.

It is not just typing in multiple places, but executing a vim command in multiple places. Which means, you can replace the insides of a parenthesis with something, then proceed to do the same action on a different-length parenthesis, without actually typing the whole command again. While not getting confused by other parentheses inside the first one, because the i( adverb handles that. Similar with quoted strings, words, sentences, paragraphs, etc.

By "repeat last command", I meant repeating a single command with ., which can do simple things like pasting a string, deleting something, toggling case, incrementing an integer, and probably more stuff I don't use. Then there are macros which can repeat multiple commands, and are actually not much harder to use than the dot.

Another thing I like about the repeating approach is that you don't have to find all the places beforehand. You can target some of them with a regex, then remember to do more of them with another regex, then the rest, one-by-one by hand.

I used to think that typing in multiple places at once was cool, but now I wouldn't go back. Not even if other editors are catching up. Are they, by the way? I remember Notepad++ could not do much more than insert the same string in multiple places at once, but I confess, I have never used VS Code, Sublime, Atom, or really anything that popped up while I was in the vim rabbit hole. Am I missing out?

> sure you can if you're comparing with Notepad++. But modern editors have multiple cursors

Notepad++ has multiple cursors. Vim has "repeat last command", which is better.

I see. Physicists face this problem with the Large Hadron Collider, and many possible hypotheses explaining its results.

Yet, I think many many nurses are needed to beat the 342 million.

I believe Pr[evidence | innocence] is p-value (or maybe one minus p-value, not sure). Statisticians use this routinely to test "innocence". It does not mean probability of innocence, but it means something.

Isn't this just hypothesis testing? [1]

Zero hypothesis: N deaths happen by chance, given a known probability distribution of patient deaths.

Alternative: There was a different probability distribution in play (apparently facilitated by a specific nurse).

P-value: 1 in 342 million, really convincing.

So, is the fallacy that somebody calls the number "probability" rather than "p-value"? Or am I getting it wrong?

[1] https://en.wikipedia.org/wiki/Statistical_hypothesis_testing

> all the money that previously went to a lot of workers instead goes to a small number of people who own these companies, and the workers will be destitute

All the money the rich make by selling their robot products to who? Each other? Certainly not to the destitute workers, those don't have money.

I agree that there is a risk that a minority will hoard the new resources for themselves, only it's not this simple.

Thanks, clear.

Though, I think my example was misunderstood. I meant, why is it a problem that there is no set meeting Russell's definition, but not a problem that there is no number meeting my definition. The example would violate an axiom if there was an axiom like:

  To every definition of a certain kind about natural numbers, there is a number which satisfies it.

Which of course would be rather useless, while the unrestricted comprehension axiom is necessary for the rest of the theory.

Maybe this is in the article which I could not read, but - why is this a paradox? What I see is that Russell tried to define a set, then found out that there is no set which would fulfill his definition. "Let igiveup's number be any prime number which is divisible by four." Why is this not a paradox, while Russell's definition is?

You can be wrong about your level of confidence just as you can be wrong about the rest of the world. Either is just a prediction made by your neural network.

> Elon Musk had 44 billion dollars and he didn't decide to share it with the world, he decided to buy Twitter with it.

That kind of is sharing his money with the world.

> "How will people be motivated to work," the Republicans would say,

Why should they work if they don't want to? You have a replicator.

> how will we motivate people to join the Army?"

If this is so, then give replicator technology to the enemy and you won't need an army any more.

Fail

You could say the same about a manager becoming an engineer

I admit I don't understand it fully either. If you want to read about it from more famous people than me, it is called "self-sampling argument". I'll try explain what I think I know.

Why some group is not empty is different from why it is me who is in that group, when I could very well be in the other one, like almost everyone. When the other group is much larger, then most potential "me"s will observe that they are in the larger, more typical group. The others will be surprised because they get the less common result, but most will not.

There are more intuitive examples than being human versus insect. Let's say, vast majority of people are not members of a royal family, so, knowing nothing else about myself, I should expect not to be, either. (Fits, afaik.) There are also some counterexamples, like that most people live elsewhere than in my town, but these are often tailored to me.

Going back to the plankton problem, I observe the equivalent of being a prince. Someone has to be, for sure, but is there a reason why me?

Now look at it from the other side. If it really was quadrillion times more probable to be born (hatched?) as a plankton, why exactly weren't you? Either tremendous luck, or something wrong with the assumption.

Makes sense to me. If every living being has 50 % chance of survival, so does every living human being, no matter what minority we are.

Not sure about the rest of Thanos's logic, of course...

Other humans != human species. Let alone specific other humans, like my family.

I know what happens to each individual human: They will die. I don't like that very much, but making human species survive doesn't seem to help. I don't really care if new, stranger humans will be born to learn about my time in history classes, and eventually die, too.