domain:tracingwoodgrains.com
jkf
2hr ago
that Army Corp biologist page that included fish gender.
Fish... ah, pretty clearly don't have gender? It can be pretty hard to even tell what sex they are without cutting them open, nevermind enquiring as to their feelings as to the roles imposed upon them by fish society.
Maybe next time the Army Corp biologists will write their report using proper scientific language; serves them right if you ask me.
wingdingspringking
2hr ago
This touches on something I've been wondering about for a while: Do all of these qualitative updates to LLMs actually translate to new use cases? In my case, the only two updates that have had any significant impact on my LLM use were the jump from ChatGpt3.5 to 4, and the increase of the context window from small to essentially limitless (yes it still has limits, but in day to day use I rarely hit them). Both of those happened in 2023. Since then, LLM tooling has become vastly better. But I struggle to think of anything that I can do now with an LLM that I couldn't have done in 2023 based purely on the quality of the LLM output.
Tacherus
2hr ago
Most progress for women's causes came from what one would broadly call the left.
I think there might be some revision in this statement. My understanding is that as late as the 60s the feminist vote was kinda up for grabs between republicans and democrats.
Tacherus
2hr ago
I think this is a new definition and as you point out a bit of a futile goal. I thought most people learned as teenagers they can’t control how they’re perceived and develop an internal sense of self but this fallacy seems to run rampant these days.
faceh
2hr ago
I might be willing to take you up on that.
Consider for a second that many Gacha games are already basically waifu simulators.
And millions of people in the U.S. already play those games. I mean, there's a lot of other games that they play too, but this is a POPULAR genre.
If we limited it to Gen Z males, we're talking 33 million guys, give or take, and 5% of that comes out to 1,650,000.
Among a generation that has already grown up using ChatGPT for everything, I would not find it hard to believe that 5% or so of them spend copious amounts of time talking to a digital AI avatar in a fairly intimate way... and don't feel weird about it either.
The audience is clearly there.
5% of ALL males is a bigger lift.
And I'm not sure where we'd pull reliable stats on these numbers either, but from the cursory amount of research I've done I've just about convinced myself that within a year, we'll see 5% or so of Gen Z folks having ongoing dialogues with personified AIs almost as much or more than their human friends.
SkoomaDentist
2hr ago
OTOH, it’s established that collections of data (such as phonebooks) can be copyrighted. None of the individual data items are under copyright, but the collection itself is.
phailyoor
2hr ago
I know canned animations off of the unity asset store are more likely, but I almost wonder if they used some of the optimus tech to do the movements. It could be AI controlled as well.
voters-eliot-azure
patently unbiased
3hr ago
I rarely come back to look back at comments, but the comment I replied to originally is also bait, when viewed from a different lens.
We're simply arguing about which problem is bigger, not whether either problem exists. Leaving my comment in response to phailyoor cuts back on the circlejerk that regulation is inherently bad. I mostly make comments like this when the circlejerk becomes unbearable.
Primaprimaprima
...something all admit only "TRUMP", and the Trump Administration, can do.
3hr ago
The problem with freeing and protecting women from men is that you must also free and protect women from women. … the propaganda about men being the real evil exists specifically to confuse these women about this issue
Huh?
SubstantialFrivolity
I'm not even supposed to be here today
3hr ago
I'm ok with Unicode to the extent that it is used to contain actual languages. But emoji can fuck right off from my text encoding, especially now that they have been hijacked for political purposes.
Didn't you just say "The decline should be observable within a few years"? Isn't that projecting to begin with?
Of course not, it's a testable prediction to confirm/disprove my priors on the present or near-term state of technology. @faceh said that a terrifying new superstimulus has entered the market that will destroy young men; I say that is ridiculous, and that in this current AI companion technology will barely move the needle at all in terms of parasocial escapism. The only way to adjudicate these wordviews is to make predictions that will either pass/fall, which necessarily involves wait time to see this technology hit mainstream society. This is something I note that AI maximalists generally decline to do, merely making breathless statements about how AI will change the world "soon". (These inevitably fail, but the AI keeps getting better at wordceling or shaperotating, which was not in dispute from most AI skeptics.)
I would be happy to make a 6 month window prediction, or a 10 year window prediction — eg "fewer than 5% of teenage boys will spend more than ten hours talking to AI girlfriends per year — but I doubt you/faceh would accept the first, and I wouldn't even remember the second prediction by the time it proves correct.
EDIT:
And if it AI doesn't become cheap / good enough, how does that affect the question of whether AI GFs / porn being superstimuli? The question seems completely unrelated to me.
The question is not AI GFs being superstimuli; it is them being significantly better, more seductive, and thus more dangerous superstimuli, which was OP's claim. There is no evidence that that's here; there's no reasonable evidence that it will shortly be here, and there I plant my flag.
sarker
It isn't happening, and if it is, it's a bad thing
3hr ago
It is definitely interesting. I don't think that this plus "Maxwell" in the name is a smoking gun, though, and that seems to be the extent of the evidence as far as I can tell.
Yeah this and other "mathematically unbiased" district-drawing algorithms often get plenty of upvotes on Hacker News, so I've seen them. My first issue with them is that they often have to choose some arbitrary optimization criteria to close the space of the problem (e.g. for this one why is that the "shortest" splitline should be chosen among all splitlines?).
My second issue, more practically, is that you'll never get state congress critters to give up even a little bit of power, let alone the power of the district-drawing pen. But red-blooded Americans love a good competition so I'd like to almost think that it would become some sort of televised spectacle where each party announces its next line like it's the NFL Draft. Maybe even some will gamble on it (or would have, if the BBB didn't hamstring gamblers). The congress critters would even get additional time in the limelight, which we all know is what they truly crave aside from receiving a greasing of the palms from industry buddies and pals.
ThisIsSin
Cainanites and Abelists
3hr ago
"Leftists don't want to emancipate women because they don't see the necessary connection between biology and womanhood!"
Leftist women don't want to emancipate women because they do see the necessary connection between gender and privilege brought on by scarcity.
Going all the way would remove that, and they have a pretty good thing going (this is why I see this kind of 'leftism' as a fundamentally conservative privilege-preserving movement at society's general expense).
The problem with freeing and protecting women from men is that you must also free and protect women from women. And until the women who want freedom understand the actual threat (and the women worthy of freedom do understand this; the propaganda about men being the real evil exists specifically to confuse these women about this issue, it's not actually intended for men) they'll make no progress in that area.
FiveHourMarathon
Wawa Nationalist
3hr ago
It depends if you think the typical consumer consider it positive or negative value that lives are ruined by the production of pornography.
So we're still projecting.
Didn't you just say "The decline should be observable within a few years"? Isn't that projecting to begin with?
And if it AI doesn't become cheap / good enough, how does that affect the question of whether AI GFs / porn being superstimuli? The question seems completely unrelated to me.
voters-eliot-azure
patently unbiased
4hr ago
If it's so inconsequential, why not follow the mundane processes of publishing why and how the change was made? That's my main issue with it. It's a canary in the coal mine for poor data integrity, which, taken in conjunction with the rest of the actions of the administration, is a huge red flag. It did not happen in an isolated context. If this was a corporate setting with financial or industrial data, heads would roll - even if the changes affect "very little".
The majority party/coalition draws the first straight line that bisects the population of the state.
See also the shortest-splitline algorithm.
The shortest-splitline algorithm for drawing N congressional districts:
(1) Start with the boundary outline of the state.
(2) Let A = ⌈N/2⌉ and B = ⌊N/2⌋.
(3) Among all possible dividing lines that split the state into two parts with population ratio A:B, choose the shortest.
(4) We now have two hemi-states, each to contain a specified number (namely A and B) of districts. Handle them recursively via the same splitting procedure.
MaiqTheTrue
Renrijra Krin
4hr ago
I’m positive that most people whining about this are not even aware of what the changes made will actually do. Reddit especially is the land of *people who freak out without bothering to find out what the changes actually do. Outside looking in, my answer would be “not much.” For the 99% of American veterans and their families using the VA, the gender column is a redundant sex column. Its deletion changes very little. For the 1% who are diagnosed as trans, noting it in the chart is probably trivial and will happen much like other medical history information.
But hatred feels so righteous, especially the pure hatred that comes from having no idea how anything actually affects anything else. It’s a deleted checkbox, and really that’s all that happened.
BurdensomeCount
—Your em dash is showing...
4hr ago
Zuck tries to humiliatingly pay his way to relevance with $300M offers to talent at other labs
This is the most blatant and open attempt at IP theft I have ever seen. Even in Quant Finance where everyone is at everyone else's throat all the time things don't get this open and base. Total lack of class from Zuckerberg.
VoxelVexillologist
Multidimensional Radical Centrist
4hr ago
but the bigger flaw would mostly fall for technical reasons due to clouds or nighttime imagery
Synthetic Aperture Radar can do some of these conditions, but isn't exactly equivalent to visible imagery. The technology exists and there are commercial providers operating satellites that acknowledge working with the US government.
magicalkittycat
4hr ago
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!
If anyone believes that there's gonna be definitive proof either way anytime soon, they're most likely wrong about it. It's been more than five years now without any account activity or additional evidence.
Short of this apparently random Mayalasian guy deciding "I'm done being AFK now for five years" or Maxwell saying "Yep that's my account, I don't know why I'm talking about a random Reddit account of mine to reporters but I am for some reason" we arent gonna get closure.
A lack of perfect closure is not proof a theory is incorrect, tons of stuff don't get perfect closure and are up in the air forever.
For all we know we will never get a definitive answer if OJ was a murderer or not, but thinking he did it is still a solid theory.
voters-eliot-azure
patently unbiased
4hr ago
My pet theory was that the easy way to solve gerrymandering would be to embrace its game-like structure, rather than try to regulate it into submission. Everyone's trying to build a system that is "fair", meanwhile games are the best way that humans have found to interact "fairly". The moves of the game:
- The majority party / coalition draw the first straight line that bisects the population of the state. To be mathematically clear: the party must choose two points on the border and a line will be drawn between them, with the requirement that roughly half the population lives in the two sectors created by the line.
- The minority party / coalition then draws a second line that either turns the map into (50% / 25% / 25%) sectors or (25% / 25% / 25% / 25%) sectors. To be mathematically clear: they must also choose two points, with at least one point on the border, but can choose the second point either on the border or on the first line that was drawn by the majority party.
- The parties alternate turns until all sectors have the required proportion of the population. The total number of moves can vary based on whether points are chosen on the border or on an already existing "line". Moves are always required to bisect an existing district, and a bisection can not bring a district below the required proportion of the population.
The obvious con here is that low-polygon districts don't map well to geographical and societal features (rivers, mountains, city limits), but I don't think that we're doing well with our current system anyway.
Also it doesn't work if you have a number of districts that isn't a power of 2.
Any change would require parties to submit to their minority, though, which will never happen - except through the courts maybe.
Anyway, emphatic agreement that FPTP is one of the roots in the tree of evil and Washington would have outlawed it in his Republic if he had foreseen its consequences.
I can agree he's wrong in that XAI is not even the first company that has developed AI GFs, and gooners have been working on it since day 0 of mildly competent LLMs. But you're wrong in calling it ridiculous. Qualitatively current technology is all that is required to have the impact he predicts, the rest is a question of training customized models, giving them access to your personal data, etc.
Do you think you'd be able to predict the exact inflection point for all the other technologies, as they were being developed? There was quite a few years between the first tittie I saw online, and the displacement of other forms of porn, for example.
More options
Context Copy link