Primaprimaprima
Bigfoot is an interdimensional being
"...Perhaps laughter will then have formed an alliance with wisdom; perhaps only 'gay science' will remain."
User ID: 342
Primaprimaprima
Bigfoot is an interdimensional being
13d ago
Kurzweil, as one of the original singularitarians, is likely banking on ASI to bail us out here.
Ilya Sutskever, co-founder of AI labs Safe Superintelligence (SSI) and OpenAI, told Reuters recently that results from scaling up pre-training - the phase of training an AI model that uses a vast amount of unlabeled data to understand language patterns and structures - have plateaued.
Sutskever is widely credited as an early advocate of achieving massive leaps in generative AI advancement through the use of more data and computing power in pre-training, which eventually created ChatGPT. Sutskever left OpenAI earlier this year to found SSI.
“The 2010s were the age of scaling, now we're back in the age of wonder and discovery once again. Everyone is looking for the next thing,” Sutskever said. “Scaling the right thing matters more now than ever.”
Sutskever declined to share more details on how his team is addressing the issue, other than saying SSI is working on an alternative approach to scaling up pre-training.
Primaprimaprima
Bigfoot is an interdimensional being
13d ago
I think we have plenty of evidence that Trump admires dictators and wants to become one and will work towards becoming one.
So why didn't he become a dictator during the first four years he was president? I've never heard a good response to this one. He was already president for four years, and yet we still have democracy. He's a known quantity.
Now it's very interesting that at least lots of his voters appear to either not mind this, or to see something else in him, and does this fact give me pause? Sometimes, but ultimately 99% of people I esteem and respect in the field of ideas/politics/philosophy oppose Trump so this makes it pretty easy for me to conclude that his supporters are the ones with faulty judgement.
Consider the following: I am a Trump supporter. Based on the above, I presume that you would thereby see my judgement as faulty. But the feeling is not mutual. I don't see your judgement to oppose Trump as incorrect; I just think you're a different type of person than me and you have different values, so of course you would think differently. You see me as faulty, whereas I just see you as different; and difference is not in itself a bad thing. Does this fact give you any pause?
An additional dimension is that 'democracy is in peril' is not only about elections. It's also about the ability of ideas to face off against one another in a somewhat mutually comprehended arena.
I think the left has had a profoundly more deleterious effect on intellectual discourse over the past 10+ years than anything Trump has ever done.
Of course people in this forum just think Dems lie more cunningly
I don't think the left is bad because they lie. In fact I don't think of them as being particularly untruthful at all, not anymore than the right is anyway. If I had to enumerate all my complaints with them, "lying" would not make the list. Rather, I think they're bad first and foremost because they can't tolerate dissent.
Primaprimaprima
Bigfoot is an interdimensional being
14d ago
life extension, replacing workers, treating disease
But all of these problems are reducible to text generation. In some sense every conceivable problem is, because solving the problem means writing out the solution, in language.
For “solving” medicine, just have the LLM print a formula for the drug you want. A lot of remote work just is text generation in a sense, but for physical labor, a sufficiently intelligent LLM would be able to accelerate progress in robotics significantly.
Whether LLMs can actually achieve these things though is an open question.
Primaprimaprima
Bigfoot is an interdimensional being
14d ago
Not a problem! I’m happy to continue the conversation at any time if you ever want to.
Primaprimaprima
Bigfoot is an interdimensional being
15d ago
Paywalled, but here's a summary from reddit:
"Some OpenAI employees who tested Orion report it achieved GPT-4-level performance after completing only 20% of its training, but the quality increase was smaller than the leap from GPT-3 to GPT-4, suggesting that traditional scaling improvements may be slowing as high-quality data becomes limited
- Orion's training involved AI-generated data from previous models like GPT-4 and reasoning models, which may lead it to reproduce some behaviors of older models
- OpenAI has created a "foundations" team to develop new methods for sustaining improvements as high-quality data supplies decrease
- Orion's advanced code-writing features could raise operating costs in OpenAI's data centers, and running models like o1, estimated at six times the cost of simpler models, adds financial pressure to further scaling
- OpenAI is finishing Orion's safety testing for a planned release early next year, which may break from the "GPT" naming convention to reflect changes in model development
“Some researchers at the company believe Orion isn’t reliably better than its predecessor in handling certain tasks, according to the employees. Orion performs better at language tasks but may not outperform previous models at tasks such as coding, according to an OpenAI employee. That could be a problem, as Orion may be more expensive for OpenAI to run in its data centers compared to other models it has recently released, one of those people said.”
This is one of several articles/posts/tweets coming out of the LLMsphere over the past couple of weeks that are renewing concerns over LLMs hitting diminishing returns.
Of course this is just speculation until OpenAI actually releases Orion (or whatever they end up calling it). And really we would need several models past Orion too to actually extrapolate a pattern. But this does fit with my subjective impression that the leap from GPT-3 to GPT-4 was not as big as the leap from GPT-2 to GPT-3, and the leap from 4 to o1 was not as big as the leap from 3 to 4. The fact that they're considering again releasing a new model without calling it GPT-5 is also telling. They know how psychologically important the "GPT-5" moniker has become at this point and they won't give that name to a model unless it really represents a major leap forward.
Primaprimaprima
Bigfoot is an interdimensional being
17d ago
I can't imagine there being another round of top-down enforced lockdowns.
I can!
But, really, your assumption would be conspiracy
Sure!
not the much simpler explanation that public health is bad when you cut funding for public health?
This phrasing makes it sound like the response to Covid arose naturally from "the facts on the ground". But the response to Covid was a political and ideological choice. We could have chosen differently. There was no direct unmediated causal link between the actual effects of Covid and the measures we took in response.
Primaprimaprima
Bigfoot is an interdimensional being
17d ago
If we get another pandemic under Trump and another round of global lockdowns, I will update heavily towards thinking that Covid was intentionally planned and the new one is too. Because that would just be a little too perfect.
Primaprimaprima
Bigfoot is an interdimensional being
18d ago
No. They’ve still won it. Reports of a pendulum swing are greatly exaggerated.
I don’t think most people voted for Trump as an explicit anti-woke vote. My impression is that the modal voter was voting mainly on the economy, with maybe a smattering of other more “procedural” considerations in play (the assassination attempts, the lawfare, etc). I don’t think wokeness was really on the radar this election cycle for most people.
Primaprimaprima
Bigfoot is an interdimensional being
19d ago
Oh sure. I agree completely. But if we’re stuck with universal suffrage, then the least we can do is require that people actually go to the damn polling place if they want their vote to be counted.
Primaprimaprima
Bigfoot is an interdimensional being
19d ago
The voting process should select for higher agency / lower time preference voters. That’s a good thing; those are the people who should have a bigger influence on politics. It’s outside the Overton window to require an IQ test as a precondition of voting, but thankfully it’s still within the Overton window to have some very simple and reasonable measures like, requiring that someone physically travel to a polling place, or requiring that someone procure a mail-in ballot for themselves. Any slight barrier to entry is better than canvassers going door to door and telling people “just sign on the dotted line, please” in order to harvest votes.
Primaprimaprima
Bigfoot is an interdimensional being
19d ago
It’s not just the fact that turnout overall was higher in 2020. It’s specifically the fact that Harris underperformed Biden by 15 million votes and Trump underperformed his 2020 results by 3 million votes. Why the massive difference for the Dems but not for Trump?
Primaprimaprima
Bigfoot is an interdimensional being
19d ago
Popular vote counts for Democrat candidates since 2008, per Wikipedia:
Obama 2008: 69,498,516
Obama 2012: 65,915,795
Clinton 2016: 65,853,514
Biden 2020: 81,283,501
Harris 2024: not yet finalized, but currently at 66,415,077
What is the steelman for why I shouldn't take Biden's anomalously high vote count in 2020 as evidence of fraud? I never looked too closely into the details of the 2020 fraud claims, and I'm sure that this issue has already been discussed at length previously, but it seems like it would be reasonable to revisit that discussion in light of Harris's vote count dropping back to be more in line with the historical average. (The votes are still being counted, but we can safely assume that her total vote count won't go too much higher than 70 million. Biden got over 10 million votes more than that.)
Trump's vote total of 74,223,975 in 2020 was also elevated compared to what would historically be expected of a Republican candidate, but it seems less anomalous in light of the 71,352,277 votes he's received so far this year. Whereas Harris has drastically unperformed Biden while simultaneously performing more in line with other Democrat candidates.
It's not exactly a formal statistical analysis, but this does somewhat increase my credence that there was substantial fraud in 2020. It just doesn't pass the smell test that there were ~15 million people who were that excited to vote for Biden, who had largely never voted Democrat before, and then they all just failed to materialize again in 2024.
Primaprimaprima
Bigfoot is an interdimensional being
19d ago
We’d be waiting potentially for weeks for an official call. So the major news networks use statistical modeling to project a winner within a reasonable timeframe.
Primaprimaprima
Bigfoot is an interdimensional being
19d ago
Fox and other major networks have called Georgia for Trump. Thank God that one’s not going to stretch out into the week (I hope). Good sign that we might get a quick resolution in the other swing states.
Primaprimaprima
Bigfoot is an interdimensional being
20d ago
It depends on how contested it is. Often we basically know who won by around midnight on election day, so about 16 hours from now. But famously in the 2000 election, no one knew who the winner was for weeks because no one could agree over who won the state of Florida.
Primaprimaprima
Bigfoot is an interdimensional being
20d ago
https://plato.stanford.edu/entries/moral-non-naturalism/
Most often, ‘non-naturalism’ denotes the metaphysical thesis that moral properties exist and are not identical with or reducible to any natural property or properties in some interesting sense of ‘natural’.
Primaprimaprima
Bigfoot is an interdimensional being
20d ago
What is "anti-empiricism"?
It's a bit difficult to define, because the idea I had in mind is a loose federation of beliefs and attitudes, and doesn't really have any specific criteria. In terms of concrete beliefs though, I would say that the core of it would be something like an openness to entities and propositions that we don't (or can't) have direct empirical confirmation of, like God, souls, ghosts, UFOs, etc. In more rarefied territory, it would be an affinity for philosophical positions like: hostility to logical positivism, belief in abstract (non-spatiotemporal) objects, belief in non-naturalistic moral facts. But independent of any concrete beliefs about the existence or non-existence of specific entities, I think it's also a psychological disposition to see things as being suffused with meaning and significance.
And is this an example of it?
Not really. Lots of people are recalcitrant in the face of new evidence when it contradicts their deeply held beliefs. That's more of a human trait than a left-right trait. But, if someone just has an overriding commitment to making sure that children have access to puberty blockers for some reason, I don't think there's anything metaphysically there that doesn't fit into a purely materialist/naturalist worldview.
"Anti-empiricism" is in no way intended to be an insult of course. I personally have a strong anti-empirical streak.
I'm still hoping to get some replies written to your recent posts on art, BTW. Keep up the great work.
Thank you! I really appreciate that.
Primaprimaprima
Bigfoot is an interdimensional being
20d ago
Tucker Carlson appearing on Joe Rogan experience speculated that UAPs are real, but are not alien visitors from other planets but supernatural in origin and always has been here.
Honestly based.
Why are apparently cooky beliefs entertained by top influencers on the right?
The contemporary American right is an accident of history, an odd amalgamation of people that hold disparate and often mutually contradictory beliefs. So generalizing about them as a group is difficult. Nonetheless I speculate that there are (at least) two factors at play here:
-
I believe there is a "religious temperament" that predisposes one to not only literal belief in the supernatural, but also anti-empiricism in general, more "speculative" modes of thinking, etc. An appreciable number of these people find themselves on the right for various reasons (the phrase "the religious right" exists for a reason).
-
Leftism is the socially dominant ideology in elite culture, so the right naturally attracts contrarians who are attracted to odd ideas for their own sake.
Primaprimaprima
Bigfoot is an interdimensional being
20d ago
I was considering writing a post on this exact topic, but I refrained, because I wasn't sure how to press my vague discontent into a substantive thesis.
I'll be happy when we can go back to weekly threads that don't contain the term "prediction market". I've always found the LW/SSC fascination with monetary betting to be somewhat crass. (I'd be inclined to speculate that "make all your beliefs pay rent" is only a few steps away from "make all people pay rent".)
It's sometimes claimed that intelligence just is pattern matching. So, trying to successfully predict a major one-off event like a presidential election could then be seen as the ultimate IQ test, the ultimate test of one's ability to reason in an uncontrolled and data-scarce environment. And I can see why that would be an intriguing challenge to take on. But pattern matching has its limits, and ultimately it just feels obvious to me that there's not much that can be said in terms of election forecasting besides "well, we'll see when we see".
Primaprimaprima
Bigfoot is an interdimensional being
22d ago
You even pulled the stupid language games trick here--"my last top-level post"
By my count, out of his last 100 comments, 23 were about Israel. I think that's a reasonable level of engagement with one topic.
throws in a completely unnecessary qualifier to determining whether you're a little too focused on something for the community's comfort.
I mean, if community comfort is the issue - not a single one of @coffee_enjoyer's posts has ever made me uncomfortable, and in fact I value his presence here highly.
Primaprimaprima
Bigfoot is an interdimensional being
23d ago
He talks about how previous generations of award-winning books were written by people who had actual practical "lived experience" of the things they were writing about (e.g. Hemingway actually fought in a war), often without having ever attended college.
I think the valorization of "lived experience" for writers and artists (which, in practice, typically means the valorization of a specific kind of experience, to the exclusion of others - traveling to distant places, exposing oneself to physical danger, etc) is misguided.
Consider this post, which linked to this graph, where people were asked how many unarmed black men they thought were killed by police in a single year. About one in five "very liberal" respondents said that the number was 10,000 or more - but the actual number is nowhere near that high. Now imagine that someone has the "lived experience" of watching their unarmed black male friend get shot by a police officer. Perhaps he hears one or two anecdotes from friends that they also knew people who had similar experiences. We can imagine that this experience might affect him greatly; we can imagine that he might start to think that this experience is more common than it really is, and he might go on to write an award-winning book about it, and this book might produce more people like those 1-in-5 Very Liberal respondents who think that police shootings of unarmed black men are much more common than they actually are. In this case, we would want his lived experience to at least be tempered by some "book learnin'". Otherwise, he might go on to write a book that was quite politically deleterious. There are some truths that can never be arrived at even with a lifetime of "lived experience" - there's no getting around the need for data, abstract reasoning, the need for knowledge of other people's experiences so you can find the common patterns.
Or consider all the things that are in principle impossible for anyone to have direct experience of. If you want to, say, write a book that deals with the historical connections between contemporary wokeism and Stalinism, or maybe the French Revolution - you're going to need to read other books for that. Eventually, historical events become so distant that no one alive could have experienced them.
Primaprimaprima
Bigfoot is an interdimensional being
24d ago
Thank you for taking the time to reply, I really do appreciate it.
I think this is an instance of the Motte of a Motte-and-Bailey that is commonly deployed in defense of every academic discipline that operates according to "humanities rules". Motte: "This is just a bunch of guys shooting the shit. Sometimes they even produce interesting things that I personally enjoy. Why do you, an outsider who doesn't even appreciate any of this, barge in and try to impose rules such as your 'epistemic standards'?"
Sorry if I didn't emphasize this enough, but I did say that you had to evaluate every work on a case by case basis. What I meant that "guys shooting the shit" is a helpful way to approach some continental works. I don't think it's the best way to approach all continental works. Derrida's Voice and Phenomenon lays out some claims and arguments in the philosophy of mind that, I think, can be phrased in simple and accessible terms. Foucault's Discipline and Punish is a pretty down to earth history of the development of the western prison system over the last several centuries. They're just, like, normal books. Nothing too mysterious going on.
I do think that Lacan goes off the reservation sometimes, to the point that the commentaries and secondary sources on his work are sometimes better than the original works themselves. But that doesn't mean that everything he wrote is bad, and it certainly doesn't mean that every work of continental philosophy is bad.
Bailey: "These people are the world authorities on philosophy. We pay them to do philosophy and all philosophers agree that they are the most influential and insightful philosophers, so we should defer to them in matters of philosophy."
"Deference" to an "authority" is a concept that's about as antithetical to philosophy as you can get. For every single claim in the history of philosophy, you can find examples of someone asserting the opposite; for every canonical philosopher, you can find another canonical philosopher who thought the first guy was an idiot. Even if you were to just focus on continental philosophy alone, I really can't emphasize enough just how fragmented it is. Philosophy just is debate and disagreement, in a way that no other field is. I mean it when I say that if you walk into an English department (which is where continental philosophers usually hang out in the US) you can find people who think that Lacan was bullshit and evil, Sartre was bullshit and evil, Derrida was bullshit and evil... who is the authority to defer to? No one would be able to agree!
Analytic philosophers I think are quite scrupulous about this, to a fault. Because of continental philosophy's greater focus on specific figures rather than isolated positions and arguments, it's more common to get people who are "fans" of one figure or another, and I acknowledge that sometimes it looks like they're treating them as an authority figure. Although I don't think that's really what's going on usually. When someone says "Marx said X" for example, it should be read as more like "X is a claim that was developed in Marx's work, so you should refer to his works if you want further justification for it" or "I believe X is true, but I didn't come up with it, Marx did", rather than "you should believe X because Marx said so". More like citing your sources, rather than an assertion of authority. Even if someone did start treating their favorite philosopher like an authority figure, they wouldn't be able to do so without major cognitive dissonance, for all the reasons mentioned above; they'd have to explain why there are a lot of other philosophers who think their favorite "authority" was wrong about everything.
I can't personally vet the psychology of everyone who talks about philosophy. Maybe there are some people who really do believe "X is true because Y said so". But, that doesn't reveal anything in particular about philosophy itself. That just reveals that that particular person is dumb and wrong. It would be like saying that data fabrication is a part of science because some scientists have fabricated data before.
As a result, there are Lacanians and Deleuzians sitting in IRBs and ethics boards and asking to be persuaded, in their terms, before I am allowed to use my funding to perform scientific experiments
I would be legitimately fascinated and highly interested if you could provide specific examples of that happening.
we defer to them in questions of what arguments are acceptable in politics and school; and ultimately they are what anchors the chains of trust and authority that we use to determine which political movements are legitimate (at risk of pulling clichés from the bingo board, the argument that the druggie who runs off with five pairs of sneakers as he torches the store is misguided but has his heart in the right place ultimately leads back, via many chains of simplification for political expediency, to some humanities tract full of "poetic language") and which ones are to be treated as threats.
I think you're overestimating the real world impact of academic philosophy here. I think it would be kinda neat if it actually did have that level of impact. But I don't think it does.
I believe (and please correct me if I'm wrong; I'm not trying to put words into your mouth) that you see a direct causal link between continental philosophy on one hand and contemporary wokeness on the other. It's a claim that I've seen repeated in various forms on TheMotte on multiple occasions, and I've always disagreed with it, for 3 main reasons:
-
Broadly speaking, I don't think that people hold the political positions they do because they read a book, or even because they talked to someone who read a book. I think people believe the things they do because they want certain things (wealth, power, various types of freedom, etc). The desire and the need for something concrete comes first, and then they look for an ideology later to justify it. So, for example, I don't think that DEI exists in the US today because of humanities academics. I think it exists because that's naturally the sort of thing that arises when you have an ascendant coalition of racial minorities and a demographically declining racial majority.
-
The "big names" of continental philosophy are not particularly woke (there's little in their work that would be recognizable as modern wokeness, anyway) and in fact I think there are resources in their work that could be a benefit to anti-wokeness.
-
Regarding the point about obscurantist language - do you think that's somehow necessary to leftist politics? If they were forced to only use simple words then the whole thing would collapse? Because I think that's clearly not true. You can easily put all the key tenets and arguments of wokeness into simple language. To reiterate the point above, woke people are woke because of the intrinsic content of the positions, not because they were hypnotized by a humanities tract.
Primaprimaprima
Bigfoot is an interdimensional being
26d ago
The rabbit from Zootopia is hot and a little uppity. She absolutely wants and deserves to get gangbanged by a bunch of foxes. No trauma involved at all!
Beauty is in the eye of the beholder!
Primaprimaprima
Bigfoot is an interdimensional being
26d ago
Defining art in the way you do is like defining marriage as a convenient way to save on rent
That's exactly what it's not! That would be a utilitarian definition of marriage. I am offering what I believe to be a thoroughly anti-utilitarian conception of art.
What is the work of art to you when it no longer has a use? When it doesn't teach you anything, when it doesn't help you do anything, when you don't gain anything from it? Do you only value it as long as it still has a functional purpose, as long as there is still some benefit to be gained (such as, knowledge of beauty and truth)? That's the real question.
Sure. But it seems like this is just bolstering my case. Yes, it is hard to turn the US into a dictatorship. That's why he wasn't able to do it in his first four years. We can extrapolate that he probably won't do it in his second four years either.
I don't disagree. Especially if we take a broad historical view. Going back not only through the religious right of the 80s and 90s, but going all the way back, through the centuries of western political thought; if we polled most people who could at all be classified as "rightist" throughout history, "tolerance of dissent" would probably not rank highly as a political virtue for most of them. And the right is no stranger to moral judgement and condemnation, certainly. I don't deny any of that.
Ultimately the only person I can speak for is myself. The views I have expressed here are not universal among "my side", although they are not wholly unique to me either.
The terms "left" and "right", although convenient, may not be the most accurate terms for our current political context. Perhaps "woke" and "anti-woke" might be better?
Oh sure. Some amount of animus towards your political opponents is natural and unavoidable. I get angry at people, I find myself wondering why they have to be such NPCs. But I think all of that is still importantly different from thinking that your opponents are evil. Evil is harder to come back from; there's less chance of redemption. It seems unclear how one could sincerely wish for there to be any space for "evil" to flourish in the world. If possible, I'd like for my opponents to have a space in the world where they can be happy and live their lives according to the principles they believe in. I just want them to do it away from me.
I don't think there's much that can be said in the abstract here without a concrete example to work through (what am I imposing on you, by what mechanism, etc).
More options
Context Copy link