These AI predictions never, ever seem to factor in how actual humans will determine what AI-generated media is successful in replacing human-ones, or if it will even be successful at all. It is all very theoretical and to me, shows a fundamental flaw in this style of "sit in a room reading papers/books and make supposedly rational conclusions about the future of the world."
A good example is: today, right now, it is a negative thing for your project to be known as AI-generated. The window of time when it was trendy and cool has largely passed. Having an obviously AI-generated header image on your blog post was cool two years ago, but now it is passé and marks you as behind the trends.
And so for the prediction that everything get swept up by an ultra-intelligent AI that subsequently replaces human-made creations, essays, writings, videos, etc., I am doubtful. Just because it will have the ability to do so doesn't mean that it will be done, or that anyone is going to care.
It seems vastly more likely to me that we'll end up with a solid way of verifying humanity – and thus an economy of attention still focused on real people – and a graveyard of AI-generated junk that no one interacts with at all.
the argument against AI taking over is we organize around symbols and narratives and are hypersensitive to waning or inferior memes, thereofre AI would need to reinvent itself as "not-AI" every time so we don't learn to categorize it as slop.
I might agree, but if there were an analogy in music, some limited variations are dominant for decades, and there are precedents where you can generate dominant memes from slop that entrains millions of minds for entire lifetimes. Pop stars are slop from an industry machine that is indistinguishable from AI, and as evidence, current AI can simulate their entire catalogs of meaning. the TV Tropes website even identifies all the elements of cultural slop people should be immune to, but there are still millions of people walking around living out characters and narratives they received from pop-slop.
there will absolutely be a long tail of people whose ontology is shaped by AI slop, just like there is a long tail of people whose ontology is shaped by music, tv, and movies today. that's as close to being swept up in an AI simulation as anything, and perhaps a lot more subtle. or maybe we'll just shake it off.
But even if a future AI becomes like this, that doesn't prevent independent writers (like gwern) from still having a unique, non-assimilated voice where they write original content. The arguments tend to be "AI will eat everything, therefore get your writing out there now" and not "this will be a big thing, but not everything."
The pop industry is a:
- machine
- which takes authentic human meaning
- and produces essentially a stochastic echo of it ("slop")
- in an optimization algorithm
- to predict the next most profitable song (the song that is most "likely")
So, this sounds an awful lot like something else that's very in vogue right now. Only it was invented in 1950 or 1960, not in 2017.
Humanity is pinning its future on the thought that we will hit intractable information-theoretic limitations which provide some sort of diminishing returns on performance before a hard takeoff, but the idea that the currently demonstrated methods are high up on some sigmoid curve does not seem at this point credible. AI models are dramatically higher performance this year than last year, and were dramatically better last year than the year before, and will probably continue to get better for the next few years.
That's sufficient to dramatically change a lot of social & economic processes, for better and for worse.
Currently the state-of-the-art is propped up with speculative investments, if those speculations turn out to be wrong enough, or social/economic changes force the capital to get allocated somewhere else, then there could be a significant period of time where access to it goes away for most of us.
We can already see small examples of this from the major model providers. They launch a mind-blowing model, get great benchmarks and press, and then either throttle access or diminish quality to control costs / resources (like Claude Sonnet 3.5 pretty quickly shifted to short, terse responses). Access to SOTA is very resource-constrained and there are a lot of scenarios I can imagine where that could get worse, not better.
Even "Today, the state of the art in is the worst it will ever be" in cryptography isn't always true, like post-spectre/meltdown. You could argue that security improved but perf definitely did not.
But that isn’t the claim being made, which is that some sort of AI god is being constructed which will develop entirely without the influence of how real human beings actually act. This to me is basically just sci-fi, and it’s frankly kind of embarrassing that it’s taken so seriously.
Yeah, but, better at _what_?
Cars are dramatically faster today than 100 years ago. But they still can't fly.
Similarly, LLMs performing better on synthetic benchmarks does not demonstrate that they will eventually become superintelligent beings that will replace humanity.
If you want to actually measure that, then these benchmarks need to start asking questions that demonstrate superintelligence: "Here is a corpus of all current research on nuclear physics, now engineer a hydrogen bomb." My guess is, we will not see much progress.
This is true only because publicly-accessible models have been severely nerfed (out of sheer panic, one assumes), making their output immediately recognizable and instantly clichéd.
Dall-E 2, for instance, was much better when it first came out, compared to the current incarnation that has obviously been tweaked to discourage generating anything that resembles contemporary artists' output, and to render everything else in annoying telltale shades of orange and blue.
Eventually better models will appear, or be leaked, and then you won't be able to tell if a given image was generated by AI or not.
If, in the future, there is a way to validate humanity (as I mentioned in my comment), then any real writers will likely use it.
Anyone that doesn't validate their humanity will be assumed to be an AI. The reaction to this may or may not be negative, but the broader point is that in this scenario, the AI won't be eating all human creations.
I mean, you'd probably get more of a vote using generative AI to spam stuff that aligns with your opinions or moving to Kenya to do low wage RHLF stuff...
Still, I like a lot of his writing. Especially the weird and niche stuff that most people don’t even stop to think about. And thanks to Gwern’s essay on the sunk costs fallacy, I ended up not getting a tattoo that I had changed my mind about. I almost got it because I had paid a deposit, but I genuinely decided I hated the idea of what I was going to get… and almost got it, but the week before I went to get the tattoo, I read that essay, and decided if small children and animals don’t fall victim to sunk costs, then neither should I! Literally - Gwern saved the skin on my back with his writing. Haha.
But I do know he created an enormous dataset of anime images used to train machine learning and generative AI models [1]. Hosting large datasets is moderately expensive - and it's full of NSFW stuff, so he's probably not having his employer or his college host it. Easy for someone on a six-figure salary, difficult for a person on $12k/year.
Also, I thought these lesswrong folks were all about "effective altruism" and "earning to give" and that stuff.
I don't like that now people might pigeonhole him a bit by thinking about his effective frugality but I do hope he gets a ton of donations (either directly or via patreon.com/gwern ) to make up for it.
EDIT: grammar
> Gwern was the first patient to successfully complete a medical transition to the gender he was originally assigned at birth... his older brother died of a Nuvigil overdose in 2001... his (rather tasteful) neck tattoo of the modafinil molecule
The only concrete things we know about gwern are that he's a world-renowned breeder of Maine Coons and that he is the sole known survivor of a transverse cerebral bifurcation.
He does have a neck tattoo, but it's actually a QR code containing the minimal weights to label MNIST at 99% accuracy.
The voice was uncanny. Simply hard to listen to, despite being realistic. I mean precisely that: it is cognitively difficult to string together meaning from that voice. (I am adjacent to the field of audio production and frequently deal with human- and machine-produced audio. The problem this podcast has with this voice is not unique.) The tonality and meaning do not support each other (this will change as children grow up with these random-tonality voices).
The conversation is excessively verbose. Oftentimes a dearth of reason gets masked by a wide vocabulary. For some audience members I expect the effort to understand the words distracts from the relationship between the words (ie, the meaning), and so it just comes across as a mashup of smart-sounding words, and the host, guest, and show gets lauded for being so intelligent. Cut through the vocabulary and occasional subtle tsks and pshaws and “I-know-more-than-I-am-saying” and you uncover a lot of banter that just does not make good sense: it is not quite correct, or not complete in its reasoning. This unreasoned conversation is fine in its own right (after all, this is how most conversation unfolds, a series of partially reasoned stabs that might lead to something meaningful), but the masking with exotic vocabulary and style is misleading and unkind. Some of these “smart-sounding” snippets are actually just dressed up dumb snippets.
Oh, how humbling and modest of him!
The voice was uncanny. Simply hard to listen to,
despite being realistic. I mean precisely that: it
is cognitively difficult to string together meaning
from that voice.
In order to protect Gwern's anonymity, I proposed
interviewing him in person, and having my friend Chris
Painter voice over his words after. This amused him
enough that he agreed.
But it did not sound fake to me, and the linked source indicates that it's a human.
Perhaps it sounds uncanny to you because it's a human reading a transcript of a conversation.... and attempting to make it sound conversational, as if he's not reading a transcript?
In addition to this, there are Lex Fridman's series of interviews with various key people from Anthropic [0], and a long discussion between Stephen Wolfram and Eliezer Yudkowsky on the theme of AI risk [1].
No! What’s wrong with Bitcoin is that it’s ugly. … It’s ugly to make your network’s security depend solely on having more brute-force computing power than your opponents, ugly to need now and in perpetuity at least half the processing power just to avoid double-spending … It’s ugly to have a hash tree that just keeps growing … It’s ugly to have a system which can’t be used offline without proxies and workarounds … It’s ugly to have a system that has to track all transactions, publicly … And even if the money supply has to be fixed (a bizarre choice and more questionable than the irreversibility of transactions), what’s with that arbitrary-looking 21 million bitcoin limit? Couldn’t it have been a rounder number or at least a power of 2? (Not that the bitcoin mining is much better, as it’s a massive give-away to early adopters. Coase’s theorem may claim it doesn’t matter how bitcoins are allocated in the long run, but such a blatant bribe to early adopters rubs against the grain. Again, ugly and inelegant.) Bitcoins can simply disappear if you send them to an invalid address. And so on.
First: actual visionary CEOs are a niche of a niche. Second: that is not how most companies work. The existence of the workforce is as important as what the company produces Third: who will buy or rent those services or products in a society where the most common economy driver (salaried work) is suddenly wiped out?
I am really bothered by these systematic thinkers whose main assumption is that the system can just be changed and morphed willy nilly as if you could completely disregard all of the societal implications.
We are surrounded by “thinkers” who are actually just glorified siloed-thinking engineers high on their own supply.
I haven't decided whether I agree with it, but I can see the thought behind it: the more mechanical work will be automated, but long-term direction setting will require more of a thoughtful hand.
That being said, in a full-automation economy like this, I imagine "AI companies" will behave very differently to human companies: they can react instantly to events, so that a change in direction can be affected in hours or days, not months or years.
Where is the data showing that more jobs get destroyed than created by technological disruption?
Without saying anything regarding the arguments for or against AI, I will address this one sentence. This quote is an example of an appeal to hypocrisy in history fallacy, a form of the tu quoque fallacy. Just because someone criticizes X and you compare it to something else (Y) from another time does not mean that the criticism of X is false. There is survivorship bias as well because we now have cars, but in reality, you could've said this same criticism against some other thing that failed, but you don't, because, well, it failed and thus we don't remember it anymore.
The core flaw in this reasoning is that just because people were wrong about one technology in the past doesn't mean current critics are wrong about a different technology now. Each technology needs to be evaluated on its own merits and risks. It's actually a form of dismissing criticism without engaging with its substance. Valid concerns about X should be evaluated based on current evidence and reasoning, not on how people historically reacted to Y or any other technology.
I'm not sure that it's acrually correct: I don't think we'll actually see "AI" actually replace work in general as a concept. Unless it can quite literally do everything and anything, there will always be something that people can do to auction their time and/or health to acquire some token of social value. It might taken generations to settle out who is the farrier who had their industry annihilated and who is the programmer who had it created. But as long as there's scarcity and ambition in the world, there'll be something there, whether it's "good work" or demeaning toil under the bootheel of a fabulously wealthy cadre of AI mill owners. And there will be scarcity as long as there's a speed of light.
Even if I'm wrong and there isn't, that's why it's called the singularity. There's no way to "see" across such an event in order to make predictions. We could equally all be in permanent infinite bliss, be tortured playthings of a mad God, extinct, or transmuted into virtually immortal energy beings or anything in between.
You might as well ask the dinosaurs whether they thought the ultimate result of the meteor would be pumpkin spice latte or an ASML machine for all the sense it makes.
Anyone claiming to be worrying over what happens after a hypothetical singularity is either engaging in intellectual self-gratification, posing or selling something somehow.
The link in this paragraph goes to a post on gwern website. This post contains various links, both internal and external. But I still failed to find one that supports claims about Newton's views on "progress".
> This offers a little twist on the “Singularity” idea: apparently people have always been able to see progress as rapid in the right time periods, and they are not wrong to! We would not be too impressed at several centuries with merely some shipbuilding improvements or a long philosophy poem written in Latin, and we are only modestly impressed by needles or printing presses.
We absolutely _are_ impressed. The concept of "rapid progress" is relative. There was rapid progress then, and there is even more rapid progress now. There is no contradiction.
Anyway, I have no idea how this interview got that many upvotes. I just wasted my time.
All you really need is a government or society that isn't conducive to technological development, either because they persecute it or because they just don't do anything to protect and encourage it (e.g. no patent system or enforceable trade secrets).
Even today, what we see is that technological progress isn't evenly distributed. Most of it comes out of the USA at the moment, a bit from Europe and China. In the past there's usually been one or two places that were clearly ahead and driving things forward, and it moves around over time.
The other thing that inspires the idea of a permanent medieval society is archaeological narratives about ancient Egypt. If you believe their chronologies (which you may not), then Egyptian society was frozen in time for thousands of years with little or no change in any respect. Not linguistic, not religious, not technological. This is unthinkable today but is what academics would have us believe really happened not so long ago.
Not discovering sources of cheap energy and other raw inputs. If you look carefully at history, every rapid period of growth was preceded by a discovery or conquest of cheap energy and resources. You need excess to grow towards the next equilibrium.
https://en.wikipedia.org/wiki/House_of_Wisdom#Destruction_by...
https://en.wikipedia.org/wiki/Library_of_Alexandria#Burning_...
https://en.wikipedia.org/wiki/Nalanda_mahavihara#Destruction...
https://en.wikipedia.org/wiki/Imperial_Library_of_Constantin...
https://en.wikipedia.org/wiki/List_of_destroyed_libraries#Hu...
Could you explain further what part of https://gwern.net/newton you thought didn't support my description of Newton's view?
I thought the large second blockquote in https://gwern.net/newton#excerpts , which very prominently links to https://www.newtonproject.ox.ac.uk/view/texts/normalized/THE... , fully justified my claims, which are taken directly from Newton's statements to his son-in-law, and also closely parallel other historical statements, like Lucretius, which I also present with clear references and specific blockquotes.
I'm a little mystified that you could describe any of this as not supporting it at all, and I'm wondering if you are looking at the wrong page or something?
Irrespective of the historical accuracy of the quote I've always felt this way in some form, having personally lived through the transition from a world where it felt like you didn't have to have an opinion on everything to one dominated by the ubiquitous presence of the Internet. Although not so much because I believe an advanced human civilization has destroyed itself in our current timeline, but because the presence of so many life-changing breakthroughs in such a short period of time to me indicates a unceasing march towards a Great Filter.
Then there isn't a downvote option if it proves poor.
https://news.ycombinator.com/item?id=21559620
Again I'm sorry for the negativity, but already at the time Gwern was held up by a certain, large, section of the community as an important influencer in AI. For me that's just a great example of how basically the vast majority of AI influencers (who vie for influence on social media, rather than research) are basically clueless about AI and CS and only have second-hand knowledge, which I guess they're good at organising and popularising, but not more than that. It's easy to be a cheer leader for the mainstream view on AI. The hard part is finding, and following, unique directions.
With apologies again for the negative slant of the comment.
This is a bit stark: there are many great knowledgeable engineers and scientists who would not get your point about a^nb^n. It's impossible to know 100% of of such a wide area as "AI and CS".
I think, engineers, yes, especially those who don't have a background in academic CS. But scientists, no, I don't think so. I don't think it's possible to be a computer scientist without knowing the difference between a regular and a super-regular language. As to knowing that a^nb^n specifically is context-free, as I suggest in the sibling comment, computer scientists who are also AI specialists would recognise a^nb^n immediately, as they would Dyck languages and Reber grammars, because those are standard tests of learnability used to demonstrate various principles, from the good old days of purely symbolic AI, to the brave new world of modern deep learning.
For example, I learned about Reber grammars for the first time when I was trying to understand LSTMs, when they were all the hype in Deep Learning, at the time I was doing my MSc in 2014. Online tutorials on coding LSTMs used Reber grammars as the dataset (because, as with other formal grammars it's easy to generate tons of strings from them and that's awfully convenient for big data approaches).
Btw that's really the difference between a computer scientist and a computer engineer: the scientist knows the theory. That's what they do to you in CS school, they drill that stuff in your head with extreme prejudice; at least the good schools do. I see this with my partner who is 10 times a better engineer than me and yet hasn't got a clue what all this Chomsky hierarhcy stuff is. But then, my partner is not trying to be an AI influencer.
totally agree that you can be a great engineer and not be familiar with it, but seems weird for an expert in the field to confidently make wrong statements about this.
Transformers can easily intellectually understand a^nb^n, even though they couldn't recognize whether an arbitrarily long string is a member of the language -- a restriction humans share!, since eventually a human, too, would lose track of the count, for a long enough string.
>> Regarding your linked comment, my takeaway is that the very theoretical task of being able to recognize an infinite language isn't very relevent to the non-formal, intuitive idea of "intelligence"
That depends on who you ask. My view is that automata are relevant to computation and that's why we study them in computer science. If we were biologists, we would study beetles. The question is whether computation , as we understand it on the basis of computer science, has anything to do with intelligence. I think it does, but that it's not the whole shebang. There is a long debate on that in AI and the cognitive sciences and the jury is still out, despite what many of the people working on LLMs seem to believe.
Yes, LLMs are bad at this. A similar example: SAT solvers can't solve the pigeonhole problem without getting into a loop
It is an exceptional case that requires "metathinking" maybe, rather than a showstopper issue
(can't seem to be able to write the grammar name, the original comment from the discussion had it)
Appreciate the diversity in the effort, but engineering is making things people can use without having to know it all. Far more interesting endeavor than being a human Google search engine.
And there's a limit to what you need to look up in a book. The limit moves further up the more you work with a certain kind of tool or study a certain kind of knowledge. I have to look up trigonometry every single time I need it because I only use it sparingly. I don't need to look up SLD-Resolution, which is my main subject. How much would Feynman need to look up when debating physics?
So when someone like Feynman talks about physics, you listen carefully because you know they know their shit and a certain kind of nerd deeply appreciates deep knowledge. When someone elbows themselves in the limelight and demands everyone treats them as an expert, but they don't know the basics, what do you conclude? I conclude that they're pretending to know a bunch of stuff they don't know.
________________
[1] ... some do. But they're students so it's OK, they're just excited to have learned so much and don't yet know how much they don't. You explain the mistake, point them to the book, and move on.
In the thread you linked, Gwern says in response to someone else that NNs excel at many complex real-world tasks even if there are some tasks where they fail but humans (or other models) succeed. You try to counter that by bringing up an example for the latter type of task? And then try to argue that this proves Gwern wrong?
Whether they said "regular grammar" or "context-free grammar" doesn't even matter, the meaning of their message is still the exact same.
Edit: sorry, I read "finite" as "infinite" :0 But n can be infinite and a^nb^n is still regular, and also context free. To be clear, the Chomskky Hierarchy of formal languages goes like this:
Finite ⊆ Regular ⊆ Context-Free ⊆ Context-Sensitive ⊆ Recursively Enumerable
That's because formal languages are identified with the automata that accept them and when an automaton accepts e.g. the Recursively Enumerable languages, then it also accepts the context-sensitive languages, and so on all the way down to the finite languages. One way to think of this is that an automaton is "powerful enough" to recognise the set of strings that make up a language.
Specifically, you can construct a finite automata to represent it.
The goal of influencers is to influence the segment of a crowd who cares about influencers. Meaning retards and manchildren looking for an external source to form consensus around.
In my country, when TV series are interviewing anonymous people they use specific visual language - pixellated face, or facing away from the camera, or face clad in shadow.
Having an actor voice the words is normal. But having an actor showing the anonymous person's face is an... unusual choice.
Downsides of anonymity: no free heroin
Something I've noticed in spending time online is that there's a "core group" of a few dozen people who seem to turn up everywhere there are interesting discussions. Gwern (who also posts here) is probably at the top of that list.
> In Internet culture, the 1% rule is a general rule of thumb pertaining to participation in an Internet community, stating that only 1% of the users of a website actively create new content, while the other 99% of the participants only lurk. Variants include the 1–9–90 rule (sometimes 90–9–1 principle or the 89:10:1 ratio),[1] which states that in a collaborative website such as a wiki, 90% of the participants of a community only consume content, 9% of the participants change or update content, and 1% of the participants add content.
He's been building software for 10 years longer than I've been alive, hopefully in a few decades I'll have gained the same breadth of technical perspective he's got.
I don't know what causes such intellectual cliques to form, perhaps it's a result of an intersection of raw intellectual power and social dynamics.
I might be reading to much into the "a" (rather than "the"), but to be clear: there is a transcript at the bottom.
>I live in the middle of nowhere. I don't travel much, or eat out, or have health insurance, or anything like that. I cook my own food. I use a free gym. There was this time when the floor of my bedroom began collapsing. It was so old that the humidity had decayed the wood. We just got a bunch of scrap wood and a joist and propped it up. If it lets in some bugs, oh well! I live like a grad student, but with better ramen. I don't mind it much since I spend all my time reading anyway.
Not sure what to think of that. On one hand, it's so impressive that gwern cares only about the intellectual pursuit. On the other hand, it's sad that society does not reward it as much as excel sheet work.
> How do you sustain yourself while writing full time?
> Gwern
> Patreon and savings. I have a Patreon which does around $900-$1000/month, and then I cover the rest with my savings. I got lucky with having some early Bitcoins and made enough to write for a long time, but not forever. So I try to spend as little as possible to make it last.
Then Dwarkesh just gets stuck on this $1k/month thing when Gwern right out of gate said that savings are being used.
Who knows how much of the savings are being used or how big of a profit he got from BTC.
He's living in a place where the floor collapsed and eats (good) ramen. If it's 12k or 20k I'm not sure it makes a meaningful difference to the narrative.
https://en.wikipedia.org/wiki/Poverty_in_the_United_States
"Among individuals living alone: 19.1% lived in poverty."
Poverty line (max yearly income) for single households: $14,580
Especially if you are doing it voluntarily, 1k a month can provide you more then enough for a comfortable life in many part of the country. More so if you can avoid car ownership/insurance and health insurance (Which gwern seems to do).
A room is only a prison cell when you're not allowed to leave.
It's a reasonable tradeoff for some circumstances.
Maybe it works for maths, physics and such, and of course it's ok to philosophize, but I think those "ivory tower" thinkers sometimes lack a certain connection to reality
Unless by reality you mean YOUR slice of the world, a sheltered place of its own.
Dwarkesh has 18 splits. https://www.dwarkeshpatel.com/i/151435243/timestamps
I got 171. So roughly 9 context discussions in one time stamp.
https://gpt3experiments.substack.com/p/a-new-chunking-approa...
Then when I saw the frankly very creepy and offputting image and voice, thinking he'd been anonymised through some AI software, thought, oh no, this kind of thing isn't going to become normal is it.
Then - plot twist - I scroll down to read the description and see that that voice is an actual human voiceover! I don't know if that makes it more or less creepy. Probably more. What a strange timeline.
Gwern is an effective altruist and his influence is largely limited to that community. It would be an exaggeration to claim that he influenced the mainstream of AI and ML researchers -- certainly Hinton, LeCun, Ng, Bengio didn't need him to do their work.
He influences the AI safety crowd, who have ironically been trying to build AGI to test their AI safety ideas. Those people are largely concentrated at Anthropic now, since the purge at OpenAI. They are poorly represented at major corporate AI labs, and cluster around places like Oxford and Cal. The EAs' safety concerns are a major reason why Anthropic has moved so much slower than its competitors, and why Dario is having trouble raising the billions he needs to keep going, despite his media blitz. They will get to AGI last, despite trying to be the good guys who are first to invent god in a bottle.
By the same token, Dwarkesh is either EA or EA adjacent. His main advertiser for this episode is Jane Street, the former employer of the world's most notorious EA, Sam Bankman-Fried as well as Caroline Ellison. Dwarkesh previously platformed his friend Leopold Aschenbrenner, who spent a year at OAI before he wrote the scare piece "Situation Report" made the rounds. Leopold is also semi-technical at best. A wordcel who gravitated to the AI narrative, which could describe many EAs.
People outside of AI and ML, please put Dwarkesh in context. He is a partisan and largely non-technical. The way he interfaces with AI is in fantasizing about how it will destroy us all, just as he and Gwern do in this interview.
It's sad to see people who are obviously above average intelligent waste so much time on this.
In what sense? Claude is way better than any other LLM available, in the sense of providing useful/correct/intelligent output.
Gwern's probably not a "wordcel" either, he can program, right? I've never seen any of his publications though.
It's called Situational Awareness too, not "Situation Report", and Yudkowsky said he didn't like it. Not that Yudkowsky is an EA either.
I think the situation is more complex than you think it is, or at least more complex than you're pretending.
Edit: oh, you're saying Gwern is an EA too? Do you have a source for that?
I didn't claim Yudkowsky was an EA. He's not. He's a rationalist who originated many thoughts about AI doom, which have influenced EAs.
Look, Gwern and Dwarkesh and Yud are all AI doomers, which is where the rationalists and the EAs overlap. Gwern is all over LessWrong, which kind of springs out of the rationalist community and has all sorts of EAs there now. But of course, the first rule of EA is the first rule of Fight Club. That's why Dario consistently denies he's an EA, even though his chief of staff was formerly running Sam Bankman-Fried's philanthropic grants.
The weird thing in these conversations between guys like Dwarkesh and Gwern is how much they wallow in the masochism of imagining a world taken over by AGI. Yud has a similar attraction to pain, but he channels it into his silly BDSM novels.
https://www.amazon.com/Dark-Lords-Answer-Eliezer-Yudkowsky-e...
If that's how they wan't to spend their time, fine. Let's all just recognize it's an acquired taste, a self-indulgence, and the furthest thing from a rational act.
He does Haskell stuff (he mentions it on his website), but he’s even better at words than he is as a mid-tier (I guess) Haskell programmer.
Whether this counts as a “wordcel” is an exercise left to the reader.
He says he has collaborators under the "Gwern" name now, but the main guy is the main guy and it's unlikely he could hide it.
How many citations for "Branwen 2018" are on the ArXiv now?
Neither mpv nor my browser can play the interview.
Edit: The YouTube link from an earlier submisison works: <https://www.youtube.com/watch?v=a42key59cZQ>.
Sad commentary when a YouTube link affords greater accessibility (I'm listening now via mpv) than a podcast URL.
Sigh.
Thing is: Are you absolutely sure that notion of human biodiversity is wrong? IQ is heritable, as height is heritable. You'll grant that there are populations that differ in their genetic potential for height -- e.g. Dalmatians vs. Pygmies -- so how is it that you dismiss out of hand the notion that there might be population-wide differences in the genetic potential for intelligence?
I can hear it now: "But IQ is not intelligence!" I agree to a point, but IQ -- and, strangely, verbal IQ in particular -- maps very neatly to one's potential for achievement in all scientific and technological fields.
The Truth is a jealous goddess: If you devote yourself to her, you must do so entirely, and take the bad along with the good. You don't get to decide what's out of bounds; no field of inquiry should be off-limits.
I'm not gonna say that people there don't think that hbd is real, but it's not an everyday discussion topic. Mostly because it's kind of boring.
(Do spend five minutes on #lesswrong! We don't bite! (Generally! If you come in like "I heard this was the HBD channel", there may be biting, er, banning.))
"In order to protect Gwern's anonymity, I proposed interviewing him in person, and having my friend Chris Painter voice over his words after. This amused him enough that he agreed."
Gwern, how did you end up with free gym? Interested for myself.
Btw, the article gets it right and starts with 'Gwern is a pseudonymous researcher and writer.'
Headlines are seldom decided by the writers.
This leads me to believe that the content and the subsequent posts are for self promotion purposes only.
The worst I can say is that I find his predictions around AI (i.e. the scaling laws) to be concerning.
edit: having now read the linked interview, I can provide a clearly non-misanthropic quote, in response to the interviewer asking gwern what kind of role he hopes to play in people's lives:
I would like people to go away having not just been entertained or gotten some useful information, but be better people, in however slight a sense. To have an aspiration that web pages could be better, that the Internet could be better: “You too could go out and read stuff! You too could have your thoughts and compile your thoughts into essays, too! You could do all this!”> You have one Steve Jobs-type at the helm, and then maybe a whole pyramid of AIs out there executing it and bringing him new proposals
Very interesting in a short story (or a side quest in Cyberpunk 2077 - yeah that one). Not so much for a description of our future.