Posts
Comments
Here's a thought experiment.
In version A, I have a button that non invasively scans my brain and creates 10 perfect copies of my brain state in a computer. I press the button. For an instant, 11 identical mind states exist in the universe. Then each mind starts diverging along different causal chains.
Intuitively, I expect the following:
- I won't experience anything unusual after pressing the button (eg, I won't wake up in a computer). I will still feel that I am in my physical body, in the room with the button
- each of the mind copies will feel that they are the 'one true version of vlad' and won't experience the other minds 'from the inside'. presumably, they will be surprised to be in a computer and not in the room?
- if I shut down the computer and kill the 10 minds, I won't experience anything unusual
In this case, I identify myself with the embodied mind.
In version B, the setup is identical except the scan is destructive. The second I press it, my physical body is destroyed.
Now, what happens to me? There's no specific reason for me to end up in one of the minds and not the others. But I cannot go to all 10 minds at the same time — I am a single mind with its own casual chain, not a collection of minds.
For instance, imagine each of the 10 minds is caused to feel a different sensation at the same time. There's nobody to feel all 10 sensations at the same time because the minds are causally isolated. Yet I cannot say that I am feeling a particular sensation and not the others.
So in version B, I still identify myself with the embodied mind, which is destroyed — hence oblivion. Conversely, what happens to the 10 minds if I delete them from the computer? Oblivion.
(This is just my attempt to map my naive intuitions. I have a sense some version of no-self could be the solution, but I'm not there yet. I also feel that naive intuitions fail for Everett branches which is another reason to be suspicious.)
Nice! Last weekend I expanded https://www.gptrim.com/ to allow the user to a) see savings in both characters and tokens; b) determine their own combination of word processing functions. Then I saw, like you said, that to save tokens you only want to remove stopwords. I will next add the option to remove punctuation. I also want to give users two general recipes: optimize for saving tokens vs. optimize for saving characters. Always happy to take more ideas.
I will probably write again on this, on my new personal Substack or other websites, reporting what I've learned. Would you like me to cite you and link to your profile? My DMs are open!
P.S: Due to my speedrunning the coding, the website now has an issue where it sometimes adds spaces to the text. I am aware of this and will fix it latest next weekend. The Python package that runs all this is accepting PRs: https://github.com/vlad-ds/gptrim.
I see your point. I think the existing tokenizer is designed to keep all parts of text, while the idea here is to sacrifice some information in favor of compression. But writing this, I also realized that this approach is more effective at saving characters than tokens.
This is what I was hoping for when I wrote this post. Thank you for your insight.
New position: sometimes when using ChatGPT, you only care about the number of characters, because of the character limit in the chat message. In that case, you want to get rid of spaces. But if you want to save on tokens, you probably should keep spaces. I think the solution is: a) allow the user to choose the mix of transformations for their use case; b) show them how much they are saving in characters and tokens so they can optimize for their use case.
Free, universal financial tracker.
I wrote an article on this subject (i.e. why do we play zero-sum games while praising positive-sum games?)
https://native-wonder.blogspot.com/2020/12/things-people-want.html
Thank you, this is very useful. Lately I've been interested in programs that are fully online and could be completed in a year. Would you have any recommendations for that?
Wonderful, thank you!
Strongly upvoted. As a Kindle-dependent newcomer who's delving into the classics, this is precious.
I have read RAZ. Does this file include it? I would actually need only the posts that are not there.
Do you plan to do this for other authors?
I had trouble understanding how the different facts and judgments in your post are connected between each other and with the concept of upside decay.
But I want to say that I really appreciate the concept, because something very similar occurred to me once, though at the time I didn't give it a name. I was studying the careers of creative artists, and there is a lot of discrimination in these fields. Against women, against people who start out in less prestigious institutions, and so on.
My idea was that because many people were excluded and diversity was stifled, this reduced the probability of "hitting the jackpot" with an extremely brilliant artist that would be the far right of the "artistic potential" curve and end up being the next Picasso. I wanted to model this intuition and verify it in the data, but eventually my project changed and I moved on. The idea, anyway, is that you reduce the chance of getting outliers (or even black swans) in the tails, but you only care about positive outliers.
Thank you for your questions, they're proving very useful.
But it is interesting to understand, what's happening to other children, who actually do math. Suddenly you realize, that "solving problems" for them is less energy demanding, which is awkward!
I'm not sure this is the case. We're humans, maths is hard for everyone. I imagine it's more about developing an ethics of work early on and being willing to delay gratification and experience unpleasant sensations for the purpose of learning something valuable. Though of course it takes a basic level of intelligence to find motivation in intellectual work. And there needs to be some specific motivation as well, i.e. math is beautiful, or math is useful.
As for the other questions... You may be getting closer than me at hitting the target here. I think the comparison between GPT-3 talk, where nothing is wrong, and "manipulation", is central.
But "manipulation" isn't like pattern-after-pattern, it is something different. What is it?
I think the whole thing revolves around mental models. Programming "clicks" when the stuff that you do with the code suddenly turns into a coherent mental model, so that you can even predict the result of an operation that you haven't tried before. I became better at programming after watching a few theoretical computer science classes, because I was more proficient at building mental models of how the different systems worked. Likewise, maths clicks when you move from applying syntactical rules to building mental models of mathematical objects.
It's easier to build mental models with programming, because the models that you're working with are instantiated on a physical support that you can interact with. And because it's harder to fool yourself and easier to get feedback. If you screw up, the computer will stop working and tell you. If you screw up with pen and paper, you might not even realize it.
This is not the whole story, but it's a bit closer to what I meant to say.
Your position is consistent, though to me somewhat troubling.
I wouldn't equate "unable to have different preferences or to envision a better situation" with "happy". Perhaps Plato's cave applies here. Or consider a child who is born in an underground prison, Banelike, and never sees the light of sun. Who is then offered the opportunity of freedom on the surface and refuses out of fear or ignorance. Would you think they are "happy"? Perhaps, but they could be happier. Or at least they could experience a richer level of existence, given that humans evolved to enjoy fresh air and nature landscapes and the feeling of the sun on their skin, something they can't even imagine at the moment.
Imagine writing a sort of will for altered-mind situations. If you fell under a hypnotism that turned you into a slave, would you want to be liberated? Or would you want people to always stop at your currently expressed preferences?
Doesn't this mean that you would connect yourself to Nozick's machine, since it would be easier to be "happy" in a state of brainwashed slavery than in the complex life of a free agent?
I agree with you, though I don't think the linked account expects an "eternal old age"; what made you think that? As I see it, it's actually an argument about the inner experience of humans and how the author thinks we wouldn't be happy with a very long lifespan. I don't agree with the author, but I linked the post as anecdotal evidence that some people who are no longer young may reject the idea of a very long lifespan because of a general feeling of life-weariness (to what extent this feeling is connected to the biological phenomenon of aging is to be ascertained).
Are you sure? That seems like a question of physics, and the accessible energy reserves and computational capacity of our light cone (the latter of which may be infinite even if the former is not).
How would computational capacity be infinite in the presence of finite energy?
Personally, I am strongly inclined towards non-interference. I have little trouble accepting that people choose wrong, knowing how fallible I am myself. I also think that, given how complex the universe is for us, it will always be easier to find arguments for inaction than for action.
And this is precisely why I am interested in arguments for interference. Most of the time, the option of non-interference is the easiest for me; which makes me at least a bit suspicious. It makes me wonder: have I carefully considered all the opposing arguments?
'Moralising' implies that I am considering interventionism in defense of my own values. I was thinking more of situations when the other is in danger or experiencing evident suffering, which evoke an empathy response.
If you saw someone fallen on the train tracks, you wouldn't shrug and say: "It's a feature of agency, let evolution work". You would try and save them. This is the kind of experience I was trying to convey.
Maybe some kind of social app inspired by liquid democracy/quadratic voting might work?
Do you think it's wise to entrust the collective with judging the worth of intellectuals? I can think of a lot of reasons this could go wrong: cognitive biases, emotional reasoning, ignorance, Dunning–Kruger effect, politically-driven decisions... Just look at what's happening now with cancel culture.
In general this connects to the problem of expertise. If even intellectuals have trouble understanding who among them is worthy of trust and respect, how could individuals alien to their field fare better?
If the rating was done between intellectuals, don't you think the whole thing would be prone to conflicts of interest, with individuals tending to support their tribe / those who can benefit them / those whose power tempts them or scares them?
I am not against the idea of rating intellectual work. I'm just mistrustful of having the rating done by other humans, with biases and agendas of their own. I would be more inclined to support objective forms of rating. Forecasts are a good example.
TIL 'former' and 'latter' are used to distinguish between two things. Corrected.
The survey is quite simplistic. 19% said "I want to live forever", while 42% said "I want to live longer than a normal lifespan, but not forever". The problem is in the ambiguity. What does 'forever' mean? A million years? Until the heat death of the universe?
And what is 'longer than a normal lifespan'? Ten years longer? A million years longer?
My guess is that most people who chose the second option want to live until they're 100 or something, and that is in fact "longer than the average lifespan" which is 79 in US.
This is confirmed by the age effect. The proportion of people who want to live forever drops from 24% to 13% from the youngest to the oldest age group. And the proportion of people who want the normal lifespan increases from 19% to 29%. But the proportion of people who want to live longer than a normal lifespan stays unchanged.
So if there is an age effect that makes anti-deathism attractive to the young but not to the old (here is a first-person account of the phenomenon) the fact that it doesn't show in the "longer than a normal lifespan" category suggests that the people in this bucket (who are the majority) are not anti-deathists. They just want to live one or two decades more than the average, and this preference stays constant throughout the life course.
Hence, I suspect that at most, only the 19% who said "I want to live forever" qualify as anti-deathists. Although some anti-deathists may have been lost to the second category because they were put off by the fact that it's impossible to literally live forever.
However, "I want to live forever" does not automatically translate in: "I would support scientific research into fighting death." It might be an idle statement that you don't intend to act upon, like when my dad says "I want a Ferrari". And what about religions or magical systems who promise immortality?
It's hard to disentangle all this from a single question. But I would argue that the proportion of genuine anti-deathists is probably lower than 19%. While the proportion of deathists is at least 60%.
The general trend has been to make computers user-friendly, and to hide the complexity from the user. On one hand, this has been helpful for their diffusion, and I'm sure it benefited a lot of people in a lot of ways (besides making a lot of money). If I think of my parents, for instance, I can't believe they would have ever started to use computers had they been more complicated.
On the other hand, this might be the fundamental obstacle in the way of coding literacy. To do stuff in the modern world, you actually have to know how to read and write. To use computers, you don't need to know how to program (at the level that most people use them). If computers keep getting more intuitive, interactive and user friendly, why should people feel the need to understand them?
(One could imagine a future where technology is so attuned to people's intentions, they can just think, gesture or say what they want, and the machine provides; as a result, they lose interest in reading and writing, and a new dark age of illiteracy begins).
To stay on computer science analogies, this reminds me of the principle of abstraction. When you call an API, it sort of feels like magic. A task gets done, and you trust that it was done correctly, and that saves you the time of controlling the code and rewriting it from scratch. "We have only to think out how this is to be done once, and forget then how it is done." (A. Turing, 1947).
Thank you for the welcome and the feedback! Yes, I am set on deepening my engagement with Reality and tackling more practical tasks. I also want to work on keeping score on my judgments and getting better at detecting & analysing my mistakes. I will definitely write more about it when the moment comes.