site banner
Advanced search parameters (with examples): "author:quadnarca", "domain:reddit.com", "over18:true"

Showing 17 of 17 results for

domain:academic.oup.com

Do you expect OnlyFans, Pornhub, VTubers, Twitch, etc to start suffering big time because a more stimulating version of the same thing has emerged? The decline should be observable within a few years.

You're going a bit far with VTubers / Twitch, as they're a bit more generic than pornography, and I don't know about the timeline, but yeah. If AI GFs / AI generated porn becomes good and cheap enough, I fully expect their human-generated variants to crash.

I mean, I think the missing ingredient is that straight men do not want to be sexually harassed by other men. They pay good money for it from female sex workers, but not from other men.

Nobody really envies cute young secretaries with creepy gross old bosses.

dating sites are more of a superstimuli than speed dating bars? Social media vs. talking to people IRL? Watching porn on VHS vs on your phone?

.... the superstimuli lies in having an interactive agent that actively adapts to your prompts, your life circumstances, etc.

So, from each of your named examples, you see one rapidly increased at the expense of the other.

Do you expect OnlyFans, Pornhub, VTubers, Twitch, etc to start suffering big time because a more stimulating version of the same thing has emerged? The decline should be observable within a few years. I on the other hand expect that all those will continue to do just fine, because they're more or equally stimulating to Grok AI companions.

I know nothing about those hospitals and next to nothing about medical financing in general, but I would estimate the likelyhood of blaming hospital closures on whatever Trump just did a week or so ago being a lie as "extremely high". It's too fast to be a sole immediate reason. Could some cuts cause some hospital that has been circling the drain for years to finally pull the plug? Sure. But the "circling the drain for years" needs to be at least as important part of the story as Trump then. If it is not, I think it'd be fair to conclude no objective inquiry of the matter is attempted and the whole story is just another "orange man bad" bullshit.

I don't see it. I don't think this is more of a superstimulus than reading/watching/playing Strawberry 100% in 2002 and imagining you're the generic high school boy they're talking to.

Is there a particular reason why dating sites are more of a superstimuli than speed dating bars? Social media vs. talking to people IRL? Watching porn on VHS vs on your phone? Doomscrolling vs. reading the paper?

For me it's pretty clear - the superstimuli lies in having an interactive agent that actively adapts to your prompts, your life circumstances, etc. Something a scripted story cannot do by definition.

She doesn't get to say "my competent engineering, which I've worked hard to develop, is the value I offer the world," because the people around her have already decided that her key value is either (a) tits or (b) decorative diversity points,

As far as I can say she does, in fact, get to say this. Literally what is standing in her way? Who will contradict her claim?

I don't see it. I don't think this is more of a superstimulus than reading/watching/playing Strawberry 100% in 2002 and imagining you're the generic high school boy they're talking to. Then streamers and camgirls emerged for the personal touch. This is just a technically impressive but less potent instantiation of what we already have.

Register my prediction as "Society reached the saturation point on pornography and parasocial escapism without AI in the early social media era". The level of social dysfunction will increase because older cohorts are dying and social mores are decaying, but I don't expect Gen Alpha will be any more goonerish than Gen Z because of this technology.

Well, I'd argue "bureaucracy" is an overly narrow conception of what the problem is with "big government."

I don't know how much "revolving door" you think there is, but it's not all that much in my experience in the DoD/IC. Mostly, people leave federal/mil service to become a contractor for more money doing much the same job.

Mostly though, the idea that you can map any given government agency onto a model where it always or by default seeks to maximize its size/budget/power/whatever is empirically false. That is often true, but it's a loose assumption. Or often various subunits of a given agency have ambitious careerists trying to maximize their impact via mission growth, but that is a zero-sum competition by default as the overall agency has a set budget.

Mostly, as someone with a (past) career and professional education in government bureaucracy, I get a bit up in arms about simplistic notions of government bureaucracy because it leads to obvious idiocy like DOGE, instead of actually getting us limited, effective government.

Uh, I haven't specifically been keeping track of most suggestions I'm afraid. I tried to go through my chat history for specific examples, but came up short since it doesn't save conversations more than a week or two old. It did note some flaws that I personally agree with, such as a predilection towards run-on sentences or instances where I'm being unclear. Most of the time, I would have run across and then fixed the flaws myself, but this approach saves me a lot of time. Unlike most authors, I spend far less time editing than writing by default. I should probably be doing more of that, and the LLMs help.

I think I get the most utility when I ask the model to rewrite whole essays for clarity, or to see how some other author would have approached that. This occasionally produced novel (to me) insights, or particular turns of phrase I might feel tempted to steal.

That is a good point.

Also his scope may not have been zeroed very well.

That's my understanding. Probably the Americans have 10 or so super-cameras hidden up there but moving the orbits of a geo-stationary satellite regularly to focus on different targets would require unsustainable amounts of propellant - I doubt they monitor anything except the highest value targets.

I imagine the hardest bit with all the low-orbit satellites is collating the data between them properly and adjusting for the differences in perspective or whatever. Night-time imagery isn't a problem I think - you can use black-body (thermal) radiation plus reflected light from human sources plus whatever weird spectra you can find floating around. But yeah, I think you'd be getting every few hours or something.

You can always make a new market with a new deadline.

In my experience, the fun thing about many people who overconfidently believe total nonsense are also overconfident that they will be proven right in short order (for current events). You'll see!

So inasmuch as Epstein Fans believe this whole case is gonna get blown wide open they also might believe it's likely the Reddit account issue will be definitively resolved in their favor.

The Foreign Service is who runs State (leaving aside the whole appointee issue). I don't know what the downsizing breakdown was. But that's not what we've been arguing.

You need to understand that monetary comp is but one thing people look for in their careers. And that many ambitious and highly capable people optimize for something other than wealth in their utility function. The IQ -> Income correlation is positive, but weaker than merely "smart people do things to make more money." Salespeople, for example, can be talented and wealthy from hard work and charisma, more than being "very capable" in the same dimensions as a biologist making far less money researching some fly.

Inasmuch as the FSOT is g-loaded at all you're getting pretty smart people into the Foreign Service. But you're also getting ideologically self-selected people. Same general issue as much of academia and teaching and government at large.

You don't want 90th percentile, you want 99.9th percentile people for your important diplomatic roles.

The funny thing about this is how much of US diplomacy is not carried out by career diplomats. Dang appointees.

What are you "transcending", and how? How do you not already have the "dignity of self-authorship"? What are you talking about? Well, let's start with the objective facts of the matter. Women can already "self-author" themselves into essentially anything. Vice President (admittedly not President of the United States yet, but there's no reason we couldn't get there in short order), professor or artist, blue collar laborer, criminal, and anything else above, below, or in between.

I don't know, she seemed pretty clear to me. Here's the key passage that answers your specific question:

Today, women are invited to succeed, but only as women; to claim rights, but only through the vocabulary of identity.

Regardless of norms in the family or on dates, earlier-wave feminists wanted to not be judged by their gender in the marketplace, in professional and political life. The idea was, as you correctly identify, for a female engineer to be perceived by her colleagues as an engineer first and not "hey, tits!... oh yeah, and I guess it's an engineer too or sth."

The author seems to be arguing that the modern left has replaced that interaction with "hey, diversity points!... oh yeah, and I guess it's an engineer too or sth." Either way, the individual woman is reduced to a passive carrier of purely instrumental value for somebody else, and (critically) not in ways she herself chosen. She doesn't get to say "my competent engineering, which I've worked hard to develop, is the value I offer the world," because the people around her have already decided that her key value is either (a) tits or (b) decorative diversity points, neither of which redound to her personal credit or are in her control. That's what I take to be her point about self-authorship still being out of reach.

Because the male body has little to no intrinsic value, it's easier for men to become a "blank slate".

Yes, this matches how I read her argument. Although re: the intrinsic value of the male body... this is something I never quite understood about the whole female-privilege "men have to be human doings, women get to be human beings" meme. If a man longs to be passively valued for the fuckable parts of his body, by people he doesn't especially want to fuck, it seems like that should be trivially achievable by hanging out in more gay men's spaces. I'd imagine a comparable range of male body types would be admired there, and pretty young men could get nearly the same mileage a pretty young woman could get. Maybe the target audience is not quite as large, but there are easily identified locales where you'd have solid odds of finding someone appreciative. In complete seriousness, when guys complain that it would be so nice to have a body with intrinsic value in others' eyes, why do they not explore the many places where this is already true?

Right now the primary obstacle is that it costs $300 a month to run.

I'm... not sure this is true. I was able to get Companions running for a couple short prompts on my phone without any active subscription. Higher usage is supposed to be locked behind SuperGrok (30 USD/month), and I did get delays on free level. SuperGrok Heavy doesn't advertise any Companion-focused capabilities, instead emphasizing the Grok 4 and Grok 4 Heavy 'supersmart' LLMs.

((Which makes sense; most workflows I can imagine are closer to a couple nVidia 4090s, rather than the nightmare-mode power that the bigger LLM models can do. It's weird to have text be more expensive than video, for once, but compare WAN local to deepseek local, and maybe it's not as goofy.))

Conversely, I think it's going to be very interestin whether Grok gets booted from the IOS store.

I am skeptical of using LLMs to critique your writing if you are already a good writer. LLMs are not good writers in my experience, and I have a hard time imagining their critiques will be helpful if you're not making basic mistakes. Can you give any examples of truly useful writing feedback you've received?

Defense contractor.