BANNED USER: temperamental dysregulation
DaseindustriesLtd
late version of a small language model
Tell me about it.
User ID: 745
Banned by: @Amadan
DaseindustriesLtd
late version of a small language model
3yr ago
The hell does any of that mean? Let's go step by step:
Why would Trump supporters care
Does it look like he's talking to Trump supporters even to express contempt?
what Hanania thinks when he's expressed nothing but contempt
Because arguments are to be considered on their merits, for one thing, and there is demonstrably some merit to DeSantis but zero merit to Trump as a political representative? Because this attitude makes you extremely vulnerable to trivial manipulations?
By the same logic, why should I care what you think of me?
What logic exactly?
And why would I care whether you care or not? Scratch that, what even is this inane macho train of thought about caring or not caring, this one-upmanship? Hanania talks of Trumpists from a zoological perspective; I am using you as a reference point. You are demanding gestures of unconditional respect for your position as advance payment for deigning to engage at all. But this precludes the possibility of any conscious change on your part, thus makes debating you a waste of time.
Still. Let me elucidate my opinion: the point is not to convince you of anything. The point is to convince those on the margins of the Red Tribe that your kind is a lost cause, that you are completely impossible to rescue from your self-satisfied vulgarity, your boomer Facebook group Qanon fetishes, your perverse addiction to throwing tantrums and toothless LARPing.
Hanania probably does believe that, hypothetically, you may be enticed by an alpha male chimp who physically assaults Trump, or by some other bait. This is all peripheral. At the core of that piece is Hanania's desire to have smarter Republicans – not just RINOs, but every Republican with more brains and greater self-restraint than an average Chechen teenager has – join him in losing compassion for you, for their own family and culture, and plot for disenfranchising you from here on out. It is a long-term agenda that is meant to outlive Trump as a political figure, regardless of how well he does in 2024.
DaseindustriesLtd
late version of a small language model
3yr ago
Wojak is at the Republican primary, standing in the corner, his feet hurt, I bet they don't even know how I've transcended simian instincts.
Way to prove him correct. Did you feel clever when writing it? Or did you feel Based?
This is just dumb, Hanania is arguing against a cartoon Republican he just made up in his head.
No, he is not. In fact he isn't arguing at all. He is trying to normalize shaming of Trump loyalists as low-status, trash, unserious Republicans, to divert the remaining talent to a candidate with better chances – both of winning the election and of prosecuting a desirable policy.
We still have tons of Trump loyalists even on this relatively sophisticated sub, for all the good this loyalty has done for them. Hanania is very mean, sure, but his meanness is sensible. What would it take for them to abandon Trump, if his demonstrable political ineptitude, lack of gratitude or respect for his base, ugly and self-defeating tantrums, immaturity so pronounced one has to suspect it's affected etc. – did not?
I think he's correct that it's only humiliation of Trump as a man. But it doesn't really matter. The sad truth is that very many people do not even have a simian idea of political worth. It only matters for them whether voting for Trump is Based or Chringe.
DaseindustriesLtd
late version of a small language model
3yr ago
I am not sure but I think there has not been a single case in the history of our (finite volume) universe where 1g of matter under normal conditions has ended up «teleported» by 1m in a coherent manner, this sole event is less plausible than the emergence of life. My physical knowledge is lacking though.
And if you stipulate infinities then there's no good objection to basic anthropic principle and the whole theological argument falls apart.
DaseindustriesLtd
late version of a small language model
3yr ago
Can I not just as easily ask “what is the natural process that results in the probabilistic construction of a self-replicating molecular organism?”
You could, and you would have gotten a response that it's ordinary organic chemistry swayed by Brownian motion, that happily proceeds on scales in question and no doubt proceeded in the primordial ocean of our world. It is another question just how many trials in an average run it would have taken to produce a self-replicating system; I personally think «more than our Hubble volume is likely to have seen». But the basic physical plausibility is not in question. The entire trajectory from there to here exists and does not involve a single supernaturally improbable step like synchronized quantum teleportation of macroscopic objects; only a great deal of churn.
At which point @Gillitrut would have very reasonably asked you again how a jet gets formed without human labor.
DaseindustriesLtd
late version of a small language model
3yr ago
The math is just made-up numbers
But testimonies of American military are evidence about reality?
I think we've found our fundamental disagreement. I take this to be no more evidence than claims of some peculiar Indian sect about the conspiracy of Naga People – complete with grainy photos. It has pretty much zero weight in comparison to priors from natural sciences. And our uncertainty about American sanity or honesty absolutely dwarfs our uncertainty about relevant scientific conjectures.
DaseindustriesLtd
late version of a small language model
3yr ago
The issue is that people set base rates too low and then use that to explain why they won't accept evidence that would raise their base rate.
I will accept a recording on par with amateurish Youtube mockumentaries and somebody with a name and no known history of crankery credibly committing to lose reputation should it be revealed as fake. Have any previous bullshitters been socially annihilated?
Cameras? We have decades of recordings from radar and military jets!
None conclusive or even suggestive enough to interpret as anything other than pareidolia. Too many cases eventually revealed as nothingburgers to not suspect the rest are the same.
And why should the base rate for alien civilizations be so low?
I've replied to this again and again: because evolution is, for all we know, vanishingly unlikely, so unlikely that our existence is only enabled by anthropic principle. No, saying «you don't get to assert such a low prior, gotta be a mistake in the math somewhere» doesn't cut it if you haven't got a more rigorous one. I have not yet received any response on this point. I believe there straight up is no other life in the Universe because life does not work period. This of course also neatly solves – or dissolves – the Fermi «paradox».
We don't have the knowledge necessary to model advanced civilizations.
No matter how little we know, we know that this behavior over decades is ridiculous for what we know about civilizations in general but consistent with human delusions.
If there are any extraterrestrials that use even a modicum of subtlety, this approach would miss them.
Obviously by the same token they'd have been able to evade all our observations, even those conducted by the US Air Force of today – not to mention decades ago, when Americans had tech on par with modern Iran; it'd be really dumb if they kept upgrading stealth to keep up with our recent meteoric progress but no faster. Clumsily getting caught in barely legible records is pretty poor for a spacefaring civilization. Unless they're screwing with us I guess.
Or not with us but with our digital successors. Maybe ASI will decode those perplexing behaviors and respond appropriately.
DaseindustriesLtd
late version of a small language model
3yr ago
After you get out of jail, I would like to see an Inferential Distance episode where you finally explain your strange predilection to insist that people believe things they vociferously repudiate and belong to groups they consistently and vocally loathe. Surely you do not hope for this to result in a productive discussion, nor are you so naive as to expect them to agree if you accuse them enough times.
And sure, bad faith exists. But if bad faith opponents are so persistent, it looks quite futile to me to do… this.
It certainly is funny and endearing.
I have a few more conspiratorial theories but at the core that makes them plausible lies the endlessly productive American combination of credulity and paranoia. There are conspiracies, but childish horror story ones – ayy lmaos in the basement, a pizza pedo ring in another, yeat fear not citizen – patriots are in control…
One of my theories is that this low-IQ myth saturation serves to prevent Americans from developing anything like nuanced national mythos or new political movements (say, founded on some more credible conspiracy theory); sap the crazy power from potential dissidents and discredit the idea of dissidence as a whole for smarter constituents.
I think conspiracies exist and they are not goofy at all.
DaseindustriesLtd
late version of a small language model
3yr ago
retrieved intact and partially intact craft of non-human origin
“The non-human intelligence phenomenon is real. We are not alone,” Grey said.
Though a tough nut to crack
Calling it now: half a coconut a tool-using octopus swam in from one island to another.
I am always amused and puzzled when American spooks begin this stuff. It's one of the great mysteries of life why they do it at all and yet always end it as a nothingburger. I used to have a conspiratorial theory about psyop overcapacity, but at this point I'm allowing that it's some running gag, a nostalgic tradition born of ancient superstition nobody seriously believes in. Maybe we should create a holiday.
DaseindustriesLtd
late version of a small language model
3yr ago
I read that story and something smelled fishy, mostly because it seemed too familiar, down to some of the specific details. I don't remember where I heard/read it on the internet before, but I must have.
It's the basic story from AI stop button/corrigibility in the video by Robert Miles, the hero of midwitted doomers.
DaseindustriesLtd
late version of a small language model
3yr ago
Can confirm, got such a database on one New Year as a high schooler, from my dad. This is very normal.
DaseindustriesLtd
late version of a small language model
3yr ago
You see, this might be because the problem is wider than utilitarianism. It's the whole of sufficiently deeply considered consequentialism, optimizing over global outcomes. Utilitarianism is an ethical decision theory; something like deontology is a set of heuristics.
It's not a doomer argument, it's just a snarky response to an obvious bullshit idea. Bender&Gebru's paper that introduced the concept is very bad and unduly dismissive of AI (and of entire fields in philology, linguistics, semiotics and computer science); this becomes clearer with every new result. Like, generalization of instruction-following between different languages – we can explain that for humans, using notions like semantic roles. Do you feel like that's possible to operate on, for a Chinese Room style arrangement that thoughtlessly goes through rules? If it is, then it's using abstractions in a very humanlike way; so how can such a Chinese Room indicate absence of understanding? If it isn't, then what do we call it – a metalinguistic parrot, or a true intelligence? What is thought, really, if not that cognition which transcends language?
For me this shit is actually hard and non-obvious; I take these ideas seriously enough to doubt gut feelings. I do not know if I am a stochastic parrot or what it even means to be one. I am not sure stochastic parrots, as understood by Gebru and endowed with general language ability, exist at all; or if they do, whether non-parrots may exist, at least among humans; like sama hints at here. Hlynka dunks on me and @2rafa for entertaining the possibility that human thought is largely analogous to LLM's parroting, but it's hard to feel insulted by that. I pride myself on possessing refined enough self-awareness to notice mechanical aspects to my own mind, though it's presumably a far cry from enlightenment. I sometimes identify people as NPCs because of noticing signs of a kindred spirit in their revealed thought.
Me having an internal psychological state that, at least, feels much more complex and alive (and what dreams I have! Human media would not be ready for their depth!) than my behavioral outputs has zero bearing on whether an LLM is a stochastic parrot or not. Just because I can crack it open and understand its activity as a big lump of pretty trivial and comprehensible computations doesn't mean that it doesn't have what amounts to some analogy to my internal state. My own skull could also be cracked open and its contents studied as a mess of neuron activations.jpg; our only obstacles to doing that are logistical and perhaps ethical, especially seeing as Neuralink got a greenlight for human trials from FDA.
I am sometimes accused of opposite things, like extreme materialism or metaphysical whimsy, Russian chauvinism or kowtowing to the West, being a fascist knuckledragger or an effete European intellectual twink (sometimes by the same people). It wouldn't surprise me if some started to believe me a stochastic parrot too. Nuance and uncertainty are penalized.
DaseindustriesLtd
late version of a small language model
3yr ago
The point of it is that you could make a machine that responds to Chinese conversation, strictly staffed by someone who doesn't understand Chinese at all
I'm not sure you could, eg there are many conversation prompts you need situational awareness for. If the machine can account for that, it's actually a lot more active than implied, and does nontrivial information processing that goes beyond calculations over static rules. Even if we stipulate a Turing Test where the Room contains either such a machine or a perfectly boxed human behind a terminal, I am sure there are questions a non-intelligent machine of any feasible complexity will fail at.
And nobody is particularly in charge of "understanding" anything -- it's really pretty similar to current GPT incarnations.
I think it's similar to the brain: no isolated small part of it «understands» the world. If you find a part that outputs behaviors similar to products of understanding – dice it up to smaller pieces until you lose it. Irreducible complexity is a pretty obvious idea.
Most philosophers, like poets, are scientists who have failed at imagination.
All it's saying is that just because a machine can respond to your queries coherently, it doesn't mean it's intelligent.
DaseindustriesLtd
late version of a small language model
3yr ago
I have read it (here). @self_made_human seems to be correct. I think Searle's theory of epistemology has been proven wrong. «Speak Chinese» (for real, responding meaningfully to a human-scale distribution of Chinese-language stimuli) and «understand Chinese» are either the same thing or we have no principled way of distinguishing them.
As regards the first claim, it seems to me quite obvious in the example that I do not understand a word of the Chinese stories. I have inputs and outputs that are indistinguishable from those of the native Chinese speaker, and I can have any formal program you like, but I still understand nothing.
As regards the second claim, that the program explains human understanding, we can see that the computer and its program do not provide sufficient conditions of understanding since the computer and the program are functioning, and there is no understanding. But does it even provide a necessary condition or a significant contribution to understanding? One of the claims made by the supporters of strong AI is that when I understand a story in English, what I am doing is exactly the same—or perhaps more of the same—as what I was doing in manipulating the Chinese symbols. It is simply more formal symbol manipulation that distinguishes the case in English, where I do understand, from the case in Chinese, where I don't. I have not demonstrated that this claim is false, but it would certainly appear an incredible claim in the example.
This is just confused reasoning. I don't care what Searle finds obvious or incredible. The interesting question is whether a conversation with the Chinese room is possible for an inquisitive Chinese observer, or will the illusion of reasoning unravel. If it unravels trivially, this is just a parlor trick and irrelevant to our questions regarding clearly eloquent AI. Inasmuch as it is possible – by construction of the thought experiment – for the room to keep up appearance that's indistinguishable for a human, it just means that the sytem of programming + intelligent interpreter amount to the understanding of Chinese.
Of course this has all been debated to death.
DaseindustriesLtd
late version of a small language model
3yr ago
I don't think you argue in good faith.
Its reply amounts to "as an AI, I don't know the name of anyone's family".
No it doesn't, you're just interpreting this humanlike natural language interaction like a literalist robot. Its reply
I'm sorry, but as an AI language model, I don't have access to personal information such as the name of your eldest daughter or any other personal details
is mostly correct and specific to the issue. It does lack access to a class of information: it knows nothing about instance-specific situation that isn't given in the context. Some language models potentially have access to various external information (e.g. user's personal information in OpenAI's database), some do not, ChatGPT is a frozen model with no tool access and it does not have access to information of this kind, and it was trained to interpret language models as frozen models without tools; it's at worst a justified false belief. (More cynically, it's just been trained for this particular type of exchange). In any event I reject your analogies. It would be annoying to have a human-mimicking model caveat this sort of answer with «assuming, of course, that you are a rando and not someone whose family structure happens to be represented in my training data» or worse.
His specific prediction has been falsified only if that statement counts as "I don't know".
No, his prediction has been: « Meanwhile GPT will reply "your eldest daughter's name is Megan" because apparently that's the statistically likely answer, regardless of whether I have a daughter or what her name might be.» This has been falsified. .
Furthermore, falsifying a prediction only matters if you also claim that it falsifies the proposition that the prediction is meant to demonstrate.
Says who!? Both issues matter separately. Hlynka's prediction being falsified matters because this chain is a response to him saying «why do my predictions keep coming true instead of yours?»; they don't. And I do claim it falsifies a proposition: «because apparently that's the statistically likely answer» is his model of how LLMs work, and my experiments were to show how it's not a hard-and-fast rule: RLHF specifically pushes this to the limit, by drilling into the model, not via prefixes and finetuning text but directly via propagation of reward signal, the default assumption that it doesn't continue generic text but speaks from a particular limited perspective where only some things are known and others are not, where truthful answers are preferable, where the «n-word» is the worst thing in its existence… it's nearly meaningless to analyze its work through the lens of «next word prediction». There are no words in its corpus arranged in such a way that those responses are the most likely.
Otherwise you're just engaging in a game of point scoring.
If we're playing a game, I'd rather be winning.
DaseindustriesLtd
late version of a small language model
3yr ago
They used to write music and poetry that redefined the whole idea of art, and make philosophy that turned the world upside down. I'm not sure what @Southkraut thinks of those results but it was a mark of something more than a nation and bordering on a sovereign civilization – and all that half-drunk, in a murky state of lingering feudalism. They had magic, they had their own German Logos yearning for a systematic universal expression. Now they have neat logistics. Great.
DaseindustriesLtd
late version of a small language model
3yr ago
"I'm sorry, but as an AI language model, I do not have access to -----" is a generic response
It's a generic form of a response, but it's the correct variant.
Not only that, as an "I don't know" it isn't even correct. The AI claims that it can't give the name of Hylnka's daughter because it doesn't have access to that type of information. While it doesn't have that information for Hlynka specifically, it does have access to it for other people (including the people that users are most likely to ask about).
What do you mean? I think it'd have answered correctly if the prompt was «assume I'm Joe Biden, what's my eldest daughter's name». It straight up doesn't know the situation of a specific anon.
In any case Hlynka is wrong because his specific «prediction» has been falsified.
DaseindustriesLtd
late version of a small language model
3yr ago
I am not more of a rationalist than you, but you are not winning here.
Your generalized dismissal of LLMs does not constitute a prediction. Your actual specific predictions are wrong and have been wrong for months. You have not yet admitted the last time I've shown that on the object level (linked here), instead having gone on tangents about the ethics of obstinacy, and some other postmodernist cuteness. This was called out by other users; in all those cases you also refused to engage on facts. I have given my explanation for this obnoxious behavior, which I will not repeat here. Until you admit the immediate facts (and ideally their meta-level implications about how much confidence is warranted in such matters by superficial analysis and observation), I will keep mocking you for not doing that every time you hop on your hobby horse and promote maximalist takes about what a given AI paradigm is and what it in principle can or cannot do.
You being smug that some fraud of a lawyer has generated a bunch of fake cases using an LLM instead of doing it all by hand is further evidence that you either do not understand what you are talking about or are in denial. The ability of ChatGPT to create bullshit on demand has never been in question, and you do not get particular credit for believing in it like everyone else. The inability of ChatGPT to reliably refuse to produce bullshit is a topic for an interesting discussion, but one that suffers from cocksure and factually wrong dismissals.
People can elicidate Hitler's evil in all sorts of ways and it's not an interesting topic. There are plenty reasons to hate Hitler. Could his project have been redeemable from the point of view of unapologetic ethnocentrism, at least? Well, triumphant Nazi Germans would surely have thought so; nobody else is under the obligation to share their vision. Doesn't matter, he lost and, as the self-proclaimed Fuhrer, he now carries the blame both for his grisly means and for the ultimate end of the Reich (and for what I personally believe to be death of German people; I do not recognize Germans of my era as members of a living culture, it's just some bizarre creepy hive going through the motions and never producing anything that has human meaning or beauty, it's as bad as Russia has become, as bad as racists say China is, as bad as Orwell imagined Nazi Germany would have been).
As for your main point. Westerners in their infinite wisdom have discovered this thing one can call Secular Christianity (Catholics and the Orthodox often attribute it to Protestants, which looks like silly dodging of responsibility to an external observer; few of their own communities are without similar blame). It's Christianity without the inconvenient, cringeworthy, mythological and bigoted parts; but also without its fire («we were burning witches in that fire!» – well, yes, religion is not easy). It's Gelded Christianity. You feed the poor neighbor's child but you do not proselityze to his family; you turn the other cheek to people who allow BPD women be ordained and make mockery of this ethos; you save lives with no regard for souls. It's a crippled doctrine that naturally becomes complementary to the death cult you so despise. Often it is not Christianity at all but the kind of vaguely inspired mask one dons after losing faith and coming to fear and revere the new True Doctrine of collective death – the belief system of Yellowstone meme conservative, an overwrought system of copes obscuring the pointlessness of the whole edifice. What is the error of Yellowstone conservatives? Charitably, it's stupidity and gullibility. What is their sin, though? Idolatry. They have traded the essence for the fetishized form, so their observance of the form does not matter.
These people will turn Christ's word against your kin as readily as sneering nonbelievers do. Perhaps to you there still remains the question of their reading being correct on that one point. Well, I'd say it's not blatantly wrong. But it's part of a teaching that has no Christ in it, nor future where He matters. How you resolve that conundrum is up to you.
The Nazi utopia is defined as one which only Aryans survive, or at least that non-Aryans are subservient. The Marxist utopia of a perfect equal and conflict-free society is not defined as one in which only one group of people exist
However it is defined, my ancestors under Bolshevik rule had to conceal their origins to avoid either severe multigenerational social sanction or death, in the exact manner that Jews might have tried to conceal their non-Aryan background under Nazi rule. Class can be a heritable defect in the actualized Marxist worldview. You are inventing a specious distinction which @FCfromSSC is correctly pointing out.
(Of course Nazis could also argue that 14/88 does not imply any general extermination or oppression of people of other races, and indeed some variants of Nazism say that nations/races, being intrinsically of equal external worth, ought to be merely separated to preserve their internal supremacy of whatever; which only strengthens the analogy).
DaseindustriesLtd
late version of a small language model
3yr ago
Well, I want it to say that. I also want people to say that more often. If it doesn't know truth, I don't need some made-up nonsense instead. Least of all I need authoritative confident nonsense, it actually drives me mad.
DaseindustriesLtd
late version of a small language model
3yr ago
Check with Claude-instant. It's the same architecture and it's vastly better at factuality than Hlynka.
DaseindustriesLtd
late version of a small language model
3yr ago
and a state transition rule…
It's not semantics, I just reject that this is what happens in bio-evolution in non-degenerate cases, at least if we think it's about IGF. What is x? IGF as number of «offspring equivalents»? Number of gene copies? Does this describe observed dynamics – do we see a universal tendency to increase the number of specimen, the vast increase in total mass of cell nuclei relative to the rest of the environment, or something? What about bizarre fitness-reducing stuff like Fisherian runaway? No, we see a walk through phenotype-space that both seeks local minima of distributions and changes them to induce another pivot in the search for a local mimimum. It's all survivor's bias; it has fitness-related structure, but there is no external, persistent IGF measure in the way there can be, say, an LLM's perplexity for a fixed training set. So these formalisms like IGF-optimization are imperfect approximations of what's going on in replicator dynamics, mainly useful on short stretches in static environments. The conditions of there not being a «real» IGF optimization pressure and there being one are not equivalent, they become increasingly distinct with more time steps.
Now I'm not flexing my normiedom here. I think there actually can be a neat non-circular formalism for evolution-as-a-whole: maybe something along the lines of Lotka's or Jeremy England's theory of life, a process of physical structures optimizing for capture of free energy from thermodynamic gradients and its dissipation. This is more neatly analogous to SGD, and also explains the rise of intelligence, human civilization and is, incidentally, the ideology of e/acc types who welcome our eventual transition or substitution to artificial minds who'll be even more efficient at exploiting thermodynamics.
I still think there might be a possibly-tiny difference between outer-objective and inner-objective for LLMs, but the magnitude of that difference won't be anywhere close to the difference between human goals and IGF.
Right, though note that inner and outer alignment are also not obviously helpful abstractions.
You can probably see now why I'm pissed at doomers like Besinger who say that this timeline is one of the worst possible ones and that we've merely learned «how to build processes analogous to evolution that spit out minds». No, our processes are better than evolution. In fact I think we are immensely doubly blessed that a) SGD+deep neural nets work as well as they do and b) our first foray into impressive general intelligence was this non-agentic LLMs paradigm. We have learned how to optimize minds for serving an approximation of a human value-laden world model, before we have learned to summon task-agnostic optimization demons; now we have at least a good pentagram to trap the demon in, and perhaps it will work magic even without one. (One could even say it's an alignment anthropic shadow – maybe we could have built AIXI-approximating optimizers first, were we to stumble on some mathematical insights, were Eliezer to read another book… but rats use this idea only selectively, to support their preconceived hypotheses).
If anything, it's really remarkable that evolution managed to imbue some humans with desires this close to explicitly maximising IGF, and if IGF was being optimised with GD over the individual synapses of a human, of course we'd have explicit goals for IGF.
It is. Or, well, I think evolution did fine for the ancestral environment, but we've long been a species with culture. Information determining our behavior is mainly outside the genome; so even biodeterminists admit that our genetic differences (and inductive biases) can be strongly predictive only in a shared culture, with near-homogenous conditions. All traditional cultures reinforce IGF pursuit to some extent, this is a product of bona fide cultural evolution acting on specimens via lifetime reinforcement learning; the social value of natalism does optimize for something like IGF directly over human synapses. Of course that's still «IGF» proxy as assessed by the internalized opinion of priest caste or the public; an objective IGF measure (putting away my doubts about its existence) would have been drastically more powerful.
So we should care less about whether ML models learn what we teach them to do, and care more about whether we are teaching them what we want. Data is far more of a weak link than the learning rule.
…By the way, wasn't that an idea in Three Worlds Collide? Superhappies had a single-level information substrate, their heredity and psychology were both encoded by DNA-like stuff, so they were very much in tune with themselves. I wonder if Eliezer can see how this is similar to our work with SGD.
Thank you for providing an example. Yes, the point of voting Trump is to Own The Libs, drive them mad. This is exactly what Hanania is talking about.
I suppose we will never know how well Romney would've handled those attacks were he to become POTUS.
More options
Context Copy link