All of Optimization Process's Comments + Replies

That all of physics was perfectly beautiful and symmetric except for hyperspace, artificial gravity, shields and a few weapon types.

Oh, this is genius. I love this.

Ahhh! Yes, this is very helpful! Thanks for the explanation.

Question: if I'm considering an isolated system (~= "the entire universe"), you say that I can swap between state-vector-format and matrix-format via

. But later, you say...

If  is uncoupled to its environment (e.g. we are studying a carefully vacuum-isolated system), then we still have to replace the old state vector picture  by a (possibly rank ) density matrix ...

But if , how could it ever be rank>1?

(Perhaps more generally: what does it mean when a state is represented as a ran... (read more)

The usual story about where rank > 1 density matrices come from is when your subsystem is entangled with an environment that you can't observe. 

The simplest example is to take a Bell state, say 

|00> + |11>  (obviously I'm ignoring normalization) and imagine you only have access to the first qubit; how should you represent this state? Precisely because it's entangled, we know that there is no |Psi> in 1-qubit space that will work. The trace method alluded to in the post is to form the (rank-1) density matrix of the Bell state, and... (read more)

That is... a very interesting and attractive way of looking at it. I'll chew on your longer post and respond there!

I have an Anki deck in which I've half-heartedly accumulated important quantities. Here are mine! (I keep them all as log10(value in kilogram/meter/second/dollar/whatever seems natural), to make multiplication easy.)

QuantityValue
Electron mass-30
Electron charge-18.8
Gravitational constant-10.2
Reduced Planck constant-34
Black body radiation peak wavelength-2.5
Mass of the earth24.8
Moon-Earth distance8.6
Earth-sun distance11.2
log10( 1 )0
log10( 2 )0.3
log10( 3 )0.5
log10( 4 )0.6
log10( 5 )0.7
log10( 6 )0.8
log10( 7 )0.85
log10( 8 )0.9
log10( 9 )0.95
Boltzmann constant-22.
... (read more)

I thank you for your effort! I am currently missing a lot of the mathematical background necessary to make that post make sense, but I will revisit it if I find myself with the motivation to learn!

This is a good point! I'll send you $20 if you send me your PayPal/Venmo/ETH/??? handle. (In my flailings, I'd stumbled upon this "fractional step" business, but I don't think I thought about it as hard as it deserved.)

How are you defining "basically equivalent"

Nyeeeh, unfortunately, sort of "I know it when I see it." It's kinda neat being able to take a fractional step of a classical elementary CA, but I'm dissatisfied because... ah, because the long-run behavior of the fractional-step operator is basically indistinguishable from the long-run behavior of ... (read more)

3Ben
Random thoughts. You can relatively simply get a global phase factor at each timestep if you want. I don;t think a global phase factor at each step really counts as meaningfully different though. Anyway, as an example of this: f(xk−1,xk,xk+1)=xk−1 So that, at each (classical) timestep every single element of the CA tape just moves one step to the right. (So any patterns of 1's and 0's just orbit the tape in circles forever, unchanging.). Its quite a boring CA, but a simple example. We can take the quantum CA that is exactly the same, but with some complex phase factor: fq(xk−1,xk,xk+1,yk)=δykxk−1exp(iθ) Where the delta function is saying "1 iff yk=xk−1 , else 0." This is exactly the same as the old classical one (everything moves on step to the right), but this time we also have a global phase factor applied to the total system. The total phase factor is exp(iθN), where N is the total number of cells on the tape. Tiny bit more interesting: fq(xk−1,xk,xk+1,yk)=δykxk−1exp(iθxk−1) Now we only gain phase factors on values of 1, so the global phase depends on the total number of 1's on the tape, rather than its length. To get proper quantum stuff we need phase factors that are not global. (IE some relative phases). I feel like this equation below is a reasonable kind of place to start, but I have run out of time for now so might return to this later. fq(xk−1,xk,xk+1,yk)=(1/√2)[δykxk−1exp(iθ)+δykxk+1exp(−iθ)]
2[comment deleted]

I was imagining the tape wraps around! (And hoping that whatever results fell out would port straightforwardly to infinite tapes.)

I've never been familiar enough with group-theory stuff to memorize the names (which, warning, also might mean that it will take you a lot of time to write a sufficiently-dumbed-down version), but the internet suggests is related to... the Minkowski metric? I would be flabbergasted to learn that something so specific-to-our-universe was relevant to this toy mathematical contraption.

1James Camacho
I wrote up my explanation as its own post here: https://www.lesswrong.com/posts/LpcEstrPpPkygzkqd/fractals-to-quasiparticles

I think compared to the literature you're using an overly restrictive and nonstandard definition of quantum cellular automata.

That makes sense! I'm searching for the simplest cellular-automaton-like thing that's still interesting to study. I may have gone too far in the "simple" direction; but I'd like to understand why this highly-restricted subset of QCAs is too simple.

Specifically, it only makes sense to me to write as a product of operators like you have if all of the terms are on spatially disjoint regions.

Hmm! That's not obvious to me; if there... (read more)

Things have coalesced near the amphitheater. When the music kicks off again, we'll go northeast to... approximately here. 47.6309473, -122.3165802 JMJM+99F Seattle, Washington

Announcement 1: I, the organizer, will be 5-10min late. Announcement 2: apparently there's some music thing happening at the amphitheater! I'll set up somewhere northeast of the amphitheater when I get there, and post more precise coordinates when I have.

$10 bounty for anybody coming / passing through Capitol Hill: pick up a blind would-be attendee outside the Zeek's Pizza by 19th and Mercer. DM me your contact information, and I'll put you in touch, and I'll pay you on your joint arrival.

Update: the library is unexpectedly closed due to staffing issues. The event is now at Fuel Coffee, one block south and across the street.

If the chance of rain is dissuading you: fear not, there's a newly constructed roof over the amphitheater!

Hey, folks! PSA: looks like there's a 50% chance of rain today. Plan A is for it to not rain; plan B is to meet in the rain.

See you soon, I hope!

Lovely! Yeah, that rhymes and scans well enough for me!

Here are my experiments; they're pretty good, but I don't count them as "reliably" scanning. So I think I'm gonna count this one as a win!

(I haven't tried testing my chess prediction yet, but here it is on ASCII-art mazes.)

I found this lens very interesting!

Upon reflection, though, I begin to be skeptical that "selection" is any different from "reward."
Consider the description of model-training:

To motivate this, let's view the above process not from the vantage point of the overall training loop but from the perspective of the model itself. For the purposes of demonstration, let's assume the model is a conscious and coherent entity. From it's perspective, the above process looks like:

  • Waking up with no memories in an environment.
  • Taking a bunch of actions.
  • Suddenly falling unco
... (read more)
3kpreid
Your brain stores memories of input and also of previous thoughts you had and the experience of taking actions. Within the “replaced with a new version” view of the time evolution of your brain (which is also the pure-functional-programming view of a process communicating with the outside world), we can say that the input it receives next iteration contains lots of information from outputs it made in the preceding iteration. But with the reinforcement learning algorithm, the previous outputs are not given as input. Rather, the previous outputs are fed to the reward function, and the reward function's output is fed to the gradient descent process, and that determines the future weights. It seems like a much noisier channel. Also, individual parts of a brain (or ordinary computer program with random access memory) can straightforwardly carry state forward that is mostly orthogonal to state in other parts (thus allowing semi-independent modules to carry out particular algorithms); it seems to me that the model cannot do that — cannot increase the bandwidth of its “train of thought while being trained” — without inventing an encoding scheme to embed that information into its performance on the desired task such that the best performers are also the ones that will think the next thought. It seems fairly implausible to me that a model would learn to execute such an internal communication system, while still outcompeting models “merely” performing the task being trained. (Disclaimer: I'm not familiar with the details of ML techniques; this is just loose abstract thinking about that particular question of whether there's actually any difference.)

I was trying to say that the move used to justify the coin flip is the same move that is rejected in other contexts

 

Ah, that's the crucial bit I was missing! Thanks for spelling it out.

Reflectively stable agents are updateless. When they make an observation, they do not limit their caring as though all the possible worlds where their observation differs do not exist.

 

This is very surprising to me! Perhaps I misunderstand what you mean by "caring," but: an agent who's made one observation is utterly unable[1] to interact with the other possible-worlds where the observation differed; and it seems crazy[1] to choose your actions based on something they can't affect; and "not choosing my actions based on X" is how I would defi... (read more)

9Scott Garrabrant
Here is a situation where you make an "observation" and can still interact with the other possible worlds. Maybe you do not want to call this an observation, but if you don't call it an observation, then true observations probably never really happen in practice. I was not trying to say that is relevant to the coin flip directly. I was trying to say that the move used to justify the coin flip is the same move that is rejected in other contexts, and so we should open to the idea of agents that refuse to make that move, and thus might not have utility.
  • Ben Garfinkel: no bounty, sorry! It's definitely arguing in a "capabilities research isn't bad" direction, but it's very specific and kind of in the weeds.
  • Barak & Edelman: I have very mixed feelings about this one, but... yeah, I think it's bounty-worthy.
  • Kaj Sotala: solid. Bounty!
  • Drexler: Bounty!
  • Olah: hrrm, no bounty, I think: it argues that a particular sort of AI research is good, but seems to concede the point that pure capabilities research is bad. ("Doesn’t [interpretability improvement] speed up capabilities? Yes, it probably does—and Chris agrees that there’s a negative component to that—but he’s willing to bet that the positives outweigh the negatives.")

Yeah, if you have a good enough mental index to pick out the relevant stuff, I'd happily take up to 3 new bounty-candidate links, even though I've mostly closed submissions! No pressure, though!

1teradimich
I can provide several links. And you choose those that are suitable. If suitable. The problem is that I retained not the most complete justifications, but the most ... certain and brief. I will try not to repeat those that are already in the answers here. Ben Goertzel Jürgen Schmidhuber Peter J.Bentley Richard Loosemore Jaron Lanier and Neil Gershenfeld Magnus Vinding and his list Tobias Baumann Brian Tomasik   Maybe Abram Demski? But he changed his mind, probably. Well, Stuart Russell. But this is a book. I can quote. There are also a large number of reasonable people who directly called themselves optimists or pointed out a relatively small probability of death from AI. But usually they did not justify this in ~ 500 words… I also recommend this book.

Thanks for the links!

  • Ben Garfinkel: sure, I'll pay out for this!
  • Katja Grace: good stuff, but previously claimed by Lao Mein.
  • Scott Aaronson: I read this as a statement of conclusions, rather than an argument.

I paid a bounty for the Shard Theory link, but this particular comment... doesn't do it for me. It's not that I think it's ill-reasoned, but it doesn't trigger my "well-reasoned argument" sensor -- it's too... speculative? Something about it just misses me, in a way that I'm having trouble identifying. Sorry!

Thanks for the collection! I wouldn't be surprised if it links to something that tickles my  sense of "high-status monkey presenting a cogent argument that AI progress is good," but didn't see any on a quick skim, and there are too many links to follow all of them; so, no bounty, sorry!

1teradimich
My fault. I should just copy separate quotes and links here.

Respectable Person: check.  Arguing against AI doomerism: check. Me subsequently thinking, "yeah, that seemed reasonable": no check, so no bounty. Sorry!

It seems weaselly to refuse a bounty based on that very subjective criterion, so, to keep myself honest, I'll post my reasoning publicly. His arguments are, roughly:

  • Intelligence is situational / human brains can't pilot octopus bodies.
    • ("Smarter than a smallpox virus" is as meaningful as "smarter than a human" -- and look what happened there.)
  • Environment affects how intelligent a given human ends up. "
... (read more)

The relevant section seems to be 26:00-32:00. In that section, I, uh... well, I perceive him as just projecting "doomerism is bad" vibes, rather than making an argument containing falsifiable assertions and logical inferences. No bounty!

Thanks for the links! Net bounty: $30. Sorry! Nearly all of them fail my admittedly-extremely-subjective "I subsequently think 'yeah, that seemed well-reasoned'" criterion.

It seems weaselly to refuse a bounty based on that very subjective criterion, so, to keep myself honest / as a costly signal of having engaged, I'll publicly post my reasoning on each. (Not posting in order to argue, but if you do convince me that I unfairly dismissed any of them, such that I should have originally awarded a bounty, I'll pay triple.)

(Re-reading this, I notice that my "re... (read more)

1Lao Mein
Thanks, I knew I was outmatched in terms of specialist knowledge, so I just used Metaphor to pull as many matching articles that sounded somewhat reasonable as possible before anyone else did. Kinda ironic the bounty was awarded for the one I actually went and found by hand. My median EV was $0, so this was a pleasant surprise.

No bounty, sorry! I've already read it quite recently. (In fact, my question linked it as an example of the sort of thing that would win a bounty. So you show good taste!)

Thanks for the link!

Respectable Person: check. Arguing against AI doomerism: check. Me subsequently thinking, "yeah, that seemed reasonable": no check, so no bounty. Sorry!

 

It seems weaselly to refuse a bounty based on that very subjective criterion, so, to keep myself honest, I'll post my reasoning publicly. If I had to point at parts that seemed unreasonable, I'd choose (a) the comparison of [X-risk from superintelligent AIs] to [X-risk from bacteria] (intelligent adversaries seem obviously vastly more worrisome to me!) and (b) "why would I... want ... (read more)

Hmm! Yeah, I guess this doesn't match the letter of the specification. I'm going to pay out anyway, though, because it matches the "high-status monkey" and "well-reasoned" criteria so well and it at least has the right vibes, which are, regrettably, kind of what I'm after.

Nice. I haven't read all of this yet, but I'll pay out based on the first 1.5 sections alone.

Thanks for the link!

Respectable Person: check. Arguing against AI doomerism: check. Me subsequently thinking, "yeah, that seemed reasonable": no check, so no bounty. Sorry!

 

It seems weaselly to refuse a bounty based on that very subjective criterion, so, to keep myself honest, I'll post my reasoning publicly. These three passages jumped out at me as things that I don't think would ever be written by a person with a model of AI that I remotely agree with:

Popper's argument implies that all thinking entities--human or not, biological or artificial--must

... (read more)
-1Tiuto
Deutsch has also written elsewhere about why he thinks AI doom is unlikely and I think his other arguments on this subject are more convincing. For me personally, he is who gives me the greatest sense of optimism for the future. Some of his strongest arguments are: 1. The creation of knowledge is fundamentally unpredictable, so having strong probabilistic beliefs about the future is misguided (If the time horizon is long enough that new knowledge can be created, of course you can have predictions about the next 5 minutes). People are prone to extrapolate negative trends into the future and forget about the unpredictable creation of knowledge. Deutsch might call AI doom a kind of Malthusianism, arguing that LWers are just extrapolating AI growth and the current state of unalignment out into the future, but are forgetting about the knowledge that is going to be created in the next years and decades. 2. He thinks that if some dangerous technology is invented, the way forward is never to halt progress, but to always advance the creation of knowledge and wealth. Deutsch argues that knowledge, the creation of wealth and our unique ability to be creative will let us humans overcome every problem that arises. He argues that the laws of physics allow any interesting problem to be solved. 3. Deutsch makes a clear distinction between persons and non-persons. For him a person is a universal explainer and a being that is creative. That makes humans fundamentally different from other animals. He argues, to create digital persons we will have to solve the philosophical problem of what personhood is and how human creativity arises. If an AI is not a person/creative universal explainer, it won't be creative and so humanity won’t have a hard time stopping it from doing something dangerous. He is certain that current ML technology won’t lead to creativity, and so won’t lead to superintelligence. 4. Once me manage to create AIs that are persons/creative universal explainers, he th
2Cleo Nardo
(1) is clearly nonsense. (2) is plausible-ish. I can certainly envisage decision theories in which cloning oneself is bad. Suppose your decision theory is "I want to maximise the amount of good I cause" and your causal model is such that the actions of your clone do not count as caused by you (because the agency of the clone "cut off" causation flowing backwards, like a valve). Then you won't want to clone yourself. Does this decision theory emerge from SGD? Idk, but it seems roughly as SGD-simple as other decision theories. Or, suppose you're worried that your clone will have different values than you. Maybe you think their values will drift. Or maybe you think your values will drift and you have a decision theory which tracks your future values. (3) is this nonsense? Maybe. I think that something like "universal intelligence" might apply to collective humanity (~1.5% likelihood) in a way that makes speed and memory not that irrelevant. More plausibly, it might be that humans are universally agentic, such that: (a) There exists some tool AI such that for all AGI, Human + Tool is at least as agentic as the AGI. (b) For all AGI, there exists some tool AI such that for all AGI, Human + Tool is at least as smart as the AGI. Overall, none of these arguments gets p(Doom)<0.01, but I think they do get p(Doom)<0.99. (p.s. I admire David Deutsch but his idiosyncratic ideology clouds his judgement. He's very pro-tech and pro-progress, and also has this Popperian mindset where the best way humans can learn is trial-and-error (which is obviously blind to existential risk).) 

I am thinking of mazes as complicated as the top one here! And few-shot is perfectly okay.

(I'd be flabbergasted if it could solve an ascii-art maze "in one step" (i.e. I present the maze in a prompt, and GPT-4 just generates a stream of tokens that shows the path through the maze). I'd accept a program that iteratively runs GPT-4 on several prompts until it considers the maze "solved," as long as it was clear that the maze-solving logic lived in GPT-4 and not the wrapper program.)

Several unimpressive tasks, with my associated P(GPT-4 can't do it):

  • 4:1 - Write limericks that reliably rhyme and scan about arbitrary topics (topics about as complex as "an animal climbing a skyscraper")
  • 12:1 - Beat me at chess (which I'm quite bad at).
    • ("GPT-4 can beat me at chess" = "Somebody can find a non-cheaty program that maps a game-history to a prompt, and maps GPT-4's output to a move, such that GPT-4 wrapped in that translation layer can beat me.")
  • 30:1 - Solve an ASCII-art maze (e.g. solve these by putting a sequence of @s from start to finish).

I... (read more)

2Archimedes
Bing chat does pretty well at the first one. Input: Write a limerick about an animal climbing a skyscraper Response: Sure, I can try to write a limerick about that. Here is one I came up with: There once was a monkey named Jack Who decided to climb a big stack He swung from each ledge But he slipped off the edge And he landed right flat on his back I hope you like it. 😊
1janus
how big/difficult do you want the ascii mazes to be? and is few-shot ok?

I'd be interested to hear thoughts on this argument for optimism that I've never seen anybody address: if we create a superintelligent AI (which will, by instrumental convergence, want to take over the world), it might rush, for fear of competition. If it waits a month, some other superintelligent AI might get developed and take over / destroy the world; so, unless there's a quick safe way for the AI to determine that it's not in a race, it might need to shoot from the hip, which might give its plans a significant chance of failure / getting caught?

Counter... (read more)

Answer by Optimization Process40

Log of my attempts so far:

  • Attempt #1: note that, for any probability p, you can compute "number of predictions you made with probability less than p that came true". If you're perfectly-calibrated, then this should be a random variable with:

      mean = sum(q for q in prediction_probs if q<p)
      variance = sum(q*(1-q) for q in prediction_probs if q<p)
    

    Let's see what this looks like if we plot it as a function of p. Let's consider three people:

    • one perfectly-calibrated (green)
    • one systematically overconfident (red) (i.e. when they say "1%" or "99%" t
... (read more)

Plot of global infant mortality rate versus time.

I donated for some nonzero X:

  • $X to johnswentworth for "Alignment By Default", which gave a surprisingly convincing argument for something I'd dismissed as so unlikely as to be not worth thinking about.
  • $2X to Daniel Kokotajlo for "Against GDP as a metric for timelines and takeoff speeds", for turning me, uh, Against GDP as a metric for timelines and takeoff speeds.
  • $2X to johnswentworth for "When Money Is Abundant, Knowledge Is The Real Wealth", which I think of often.
  • $10X to Microcovid.org, which has provided me many times that much value.

My attempted condensation, in case it helps future generations (or in case somebody wants to set me straight): here's my understanding of the "pay $0.50 to win $1.10 if you correctly guess the next flip of a coin that's weighted either 40% or 60% Heads" game:

  • You, a traditional Bayesian, say, "My priors are 50/50 on which bias the coin has. So, I'm playing this single-player 'game':

    "I see that my highest-EV option is to play, betting on either H or T, doesn't matter."

  • Perry says, "I'm playing this zero-sum multi-player game, where my 'Knightian uncerta

... (read more)

I regret to report that I goofed the scheduling, and will be out of town, but @Orborde will be there to run the show! Sorry to miss you. Next time!

2lsusr
No big deal. I appreciate you making this happen.

you say that IVF costs $12k and surrogacy costs $100k, but also that surrogacy is only $20k more than IVF? That doesn't add up to me.

Ah, yes, this threw me too! I think @weft is right that (a) I wasn't accounting for multiple cycles of IVF being necessary, and (b) medical expenses etc. are part of the $100k surrogacy figure.

sperm/egg donation are usually you getting paid to give those things

Thanks for revealing that I wrote this ambiguously! The figures in the book are for receiving donated eggs/sperm. (Get inseminated for $355, get an egg implanted in you for $10k.)

Ooh, you raise a good point, Caplan gives $12k as the per-cycle cost of IVF, which I failed to factor in. I will edit that in. Thank you for your data!

And you're right that medical expenses are part of the gap: the book says the "$100k" figure for surrogacy includes medical expenses (which you'd have to pay anyway) and "miscellaneous" (which... ???).

So, if we stick with the book's "$12k per cycle" figure, times an average of maybe 2 cycles, that gives $24k, which still leaves a $56k gap to be explained. Conceivably, medical expenses and "miscellaneous" could fill that gap? I'm sure you know better than I!

2weft
I'm saying it's $25k PER CYCLE. (granted, this is Bay Area prices, but still) IVF requires multiple other expenses that aren't the fertilization itself. These other expenses include about $5-6k of injectable drugs that stimulate egg production, and about $6000 for the implantation.

Everything in the OP matches my memory / my notes, within the level of noise I would expect from my memory / my notes.

That's a great point! My rough model is that I'll probably live 60 more years, and the last ~20 years will be ~50% degraded, so by 60 remaining life-years are only 50 QALYs. But... as you point out, on the other hand, my time might be worth more in 10 years, because I'll have more metis, or something. Hmm.

(Another factor: if your model is that awesome life-extension tech / friendly AI will come before the end of your natural life, then dying young is a tragedy, since it means you'll miss the Rapture; in which case, 1 micromort should perhaps be feared many times more than this simple model suggests. I... haven't figured out how to feel about this small-probability-of-astronomical-payoff sort of argument.)

Load More