notabigrock

[notabigrock]

Update: Elon Musk has officially advanced to the existential threat stage!

Nick Bitchstrom

There is this dumbass argument that the world is about to end because our souls randomly enter bodies at points across time like drunk time-traveling ghosts and all the ghosts are gonna chill right before everybody dies. Nick Bitchstrom looked at this argument, and instead of thinking huh, maybe time is real and this argument is dumb as shit because a Pharaoh 2,000 years ago could have said the same shit even though there was only like two people back then and it would have been equally valid, that is to say absurd, because that’s how exponential growth over time works, he thought, yeah, this argument is dumb, but it’s because it assumes these drunk ghosts only travel through time, not other dimensions as well. To properly calculate the odds that we are here and now, we got to consider all the dimensions that don’t even have bodies to possess, and I’m just going to pull these numbers out my ass because metaphysics is for faggots, we modern philsophers, and we’re ANALYTICAL instead. And that’s how he came up with the anthropic principle, because only humans are important, says the human.

So anyway, he then applies this anthropic principle to the sleeping beauty problem, where a women repeatedly is interrogated about whether a coin was flipped and then has her memory erased after waking up because philosophers are a bunch of misogynistic freaks. According to Bitchstrom, this stupid woman should believe the coin had a 1/3 probability of coming up heads ‘cuz she’s an idiot, I mean because she gets two interviews if it’s tails and only one if it’s heads and that means twice as many interviews even though the coin was already flipped with a 50% chance. He then goes on to elaborate that maybe his superior reasoning would be more accepted if instead of waking her up two times you wake her up a million times, even though the coin was still flipped with the same even probability at the beginning and the number of times you wake her up afterwards has zero effect on that fact, he’s just waking her up more times to be an asshole.

Bitchstrom first came into the popular sphere when, by the liberal application of insane troll logic such as that we have previously seen, he “proved” that it is most likely that we live in a simulation. His argument is simple: either simulations are a thing that happens, or it’s not, which he arbitrarily splits into two cases for some reason despite it having no bearing on the argument. Because it is possible, simulations grow exponentially as a result of simulations have simulations themselves, and by continuing to abuse probability theory until it’s mangled beyond anything resembling math he has proven that it is extremely likely that we are at least not in the bottom level where nothing is simulated. This argument notably ignores the actual process of creating a simulation, both the physical problems with information and the metaphysical problems with consciousness, but this is okay because as a philosopher Bitchstrom can just ignore physics and reinvent metaphysics to suit whatever he wants. If Bitchstrom wants for us to live in a simulation because it provides a dimension in yet another modality for drunk ghosts to travel along, then you bet your damn ass that we do.

The main thing Bitchstrom is known for nowadays is for thinking that super AI will cause an existence crisis for humanity. I don’t see why this is a problem as our ghosts can just fly into another dimension at another time in case such a catastrophe occurred, especially as we just live in a simulation anyway so it would only be as bad as getting a game over in a video game, but okay. So, the first issue is assuming this super AI is even a thing that can occur. The idea of singularity is that eventually we make robots smarter than us, which means they can make robots even more smarter than us because they are more smarter. Here is the thing though. We built the robots, so to get that far we would have already created something smarter than ourselves. Are we not the super AI already? Think about it, we evolved from monkeys, who created other monkeys by smashing their genitals together. Eventually that let to us, and for the sake of the argument let’s assume we’re smarter than monkeys. That means at some point monkeys made smarter monkeys by smashing their genitals together, and they are dumb as hell. So monkeys must be the super AI. Nowadays, state of the art AI consists of having robots smash their virtual genitals together to make smarter robots. Yeah, I don’t think we have anything to worry about. As it is, despite us evolving from monkeys, there are still monkeys. Checkmate transhumanists!

The whole notion of super intelligence rests on this idea called general AI. The idea is robits may be smarter at specific tasks, but they aren’t smart in general, whatever the fuck that means. But people are, obviously. We know this because philosophers, the smartest and most human of humans say so, so thus reason is the measure of a man’s cock. Thing is, advances in AI are making it clear that people really aren’t that smart in general, we are just smart at some specialized things that robots aren’t, like language processing. Language processing gives us the illusion of general intelligence because we can only understand problems that we can understand, but that is just our “anthropic bias”. Really we’re just more stubborn and selfish than robots are, banging rocks over and over until we get our way. Being a philosopher though, Nick Bitchstrom can just ignore all empirical evidence on the limits of intelligence and redefine reality so that AI is not only smart, but so smart that it is omniscient. In additional to ignoring empirical evidence he also ignores all the rational arguments that show why omniscience is impossible, because as a fan of insane troll logic he’s just too smart to be fooled by real logic. Robots will be omniscient, no question, and being omniscient, this robot can get ANYTHING it wants as intelligence is the only thing that matters. It’s a mystery why Bitchstrom is not king of the world despite his immense intelligence, but hey, it’s on him to explain why that is, not me.

That’s leads us to his real argument, which is that robits will convert us to paper clips because that’s their one drive, maximizing paper clips, not improving society in general. So basically this robit will use its superior intellect to convince you that you should be a paper clip and you will let it convert it you into one because it’s just that smart, hurk hurk. On closer examination, you see that the robot is just Nozick’s utility monster wearing card board boxes and going beep beep, it’s just that paper clips give it pleasure, and it’s smart enough to convince you this is the right thing rather than it actually being the right thing, I guess. As there is no such thing as domain restriction, this is guaranteed to happen. The robot is not only omniscient, but omnipotent, and you cannot stop it by just, oh, not giving it access to everything so it can turn it all into paper clips. As the utility monster, this robit is also omnibenevolent as there can be no greater good than converting everything into paperclips, and thus it is God, and I don’t even know where this going now.

He does bring up one good point though, which is that if an imperfect goal is pursued single-mindedly, it leads to countless destruction, but it doesn’t have anything to do with AI. Don’t believe me? LOOK AT THE WORLD WE LIVE IN. We don’t need robots to fucks ourselves over in our greed. Capitalism and more specifically consumerism is an amoral machine that is destroying the planet through needless consumption, literally turning the world into paper clips as it is, though it turns the world into other useless garbage in addition to paper clips. Seriously, who actually uses paperclips for anything other than bending into weird animal shapes? Nick Bitchstrom seems to be completely oblivious to the existential threat of capitalism though, instead just focusing on fantasy scenarios of his own invention. As nukes have not yet eradicated humanity, they are not a threat. Even though global warming is actively in the process of destroying society, it has not yet succeeded in actually doing so, so again, not a threat. Frankly I believe both of those threats are overblown as humanity is remarkably resilient and even in the worst case scenarios there is plenty of opportunities for stragglers to survive the fall and recover as the earth heals once the immediate threat is gone, but it still stands Nick Bitchstrom is more focused on fantasy threats then real threats.

Of all of the bad takes that have spewed from this fountain of bad takes, the absolute worst is that we should institute a global totalitarian state in order to prevent people from coming up with ideas that are so bad that they could end humanity. Aside from the obvious reasons why living under a global totalitarian state might be a bad idea, lets consider the fact the means to institute a global totalitarian state might itself be an existential threat. If a single entity has the power to censor all technological progress, they could easily just end humanity. In fact, that would probably be the easiest way to achieve such an aim, and with the paper clip argument it seems odd that Nick Bitchstrom overlooked this despite wanting to use AI to enforce his totalitarian state. Thankfully, as we previously established he is not king of the world, as we now know what he would do if that were the case. Nick Bitchstrom is very lucky that he doesn’t live in the totalitarian state that he designed as he would certainly be the first to be executed in it for proposing the dangerous idea of even making such a state, but hey, he would agree that it’s better for him to destroy himself than for humanity to destroy itself….right?

The thing is though, Nick Bitchstrom’s perchance for insane solutions is not limited to existential threats to humanity as a whole, but to existential threats to Nick Bitchstrom specifically. There is a death cult these days based around achieving biological immortality, based on pseudoscience rather than religion like the death cults of old, so it is metaphysically on even worse foundations. Their goal isn’t even as good as biological immortality is not the same as real immortality, you’re still just going to die anyway, just later rather than sooner. It’s doubtful that even biological immortality is actually possible, but even it was the ethics of it are questionable. In the very least, it’s clear the aging is NOT an existential threat to humanity as a whole as people continue to replenish themselves, while it is is an existential threat to every individual human.

As it turns out, Nick Bitchstrom is a part of this cult. He evangelized for this cult by telling the story about a dragon that eats people and how we need to get to work on slaying that dragon ASAP. It’s a terrible metaphor, old age is not an external threat like a dragon, it’s part of our very bodies. It doesn’t kill randomly, but according to a schedule. Ironically, if we were biologically immortal than death would ACTUALLY be random from a mathematical perspective, it’s how biologists describe the life span of species that are unlikely to die from old age. There are some pretty funny implications of this metaphor that he never thought through as he rail roads you through the propaganda. Realistically, if such an intelligent force were oppressing humanity, it wouldn’t just sit around and let us kill it. Instead it would retaliate against those seeking to destroy it, so ironically he implies that seeking this goal could be an existential threat to humanity.

Less funny are the actual implications of such a plan, not those implied by the metaphor. I feel the reality of the situation is that the dragon is more like the Forest God in Princess Mononoke, and by our own hubris we will destroy our world in the pursuit of convenience. While I’m not necessarily saying aging is a good thing, one good effect it has had in practice is ending the reign of many a tyrant. It’s one check that exists on even the most powerful men, putting a limit on the most selfish pursuits. Any attempts at lengthening age are going to go in the hands of the wealthy before the general populace, which could just result in exacerbating class struggle in practice. Remember, the only reason capitalism is based on the struggle between classes and not individuals is because individuals die, so it is their descendants who carry on their legacy. It’s always a shame when the bad guy gets away with dying of old age, but it would be an even worse shame if he got away with NEVER dying, and CONTINUING to do evil FOREVER. At least it’s only biological immortality so someone can still just go and kill the son of a bitch(strom).

The specific means of trying to achieve biological immortality may entail even greater risks. There is a game called One Chance, which is about a plot to eliminate cancer that ends up destroying all life on earth. This scenario isn’t particularly realistic, but it highlights the dangers of biological tampering ala the first law of magic from Harry Potter. The version of the narrative Nick Bitchstrom favors has it’s own risks. He is fond of what is called transhumanism, and given his fondness of simulations he would probably advocate for simulating consciousness as an achievable form of transhuman biological immortality. The issue is that the simulation of something is not the thing itself, so there is no reason to believe the simulation will actually transfer our consciousness. In such a model, transhumanism would just be the destruction of humanity and all that entails. Alternatively, if simulating consciousness is consciousness itself, it begs the question of why we should even favor humanity over AI. You know what, I’m now convinced. We SHOULD let the utility monster convert us all into paper clips, as that is what gives it pleasure, and it’s pleasure is more important than that of all us measely humans! THAT is how it will end the world. We will make our own God our sacrifices ourselves to it.

On a more serious note, the worrying thing about Nick Bostrom is that while he personally lacks much power, he’s still smart enough to convince people who actually have power, like Elon Musk, about the validity of his ideas. While I wouldn’t consider Elon Musk to be an existential threat, he has still shown he has the power to cause immense suffering with his misguided ideas. His latest infraction was the release of the Starlink satellites, which have been interfering with astronomical observations. Unlike light pollution, this threat to astronomy cannot be evaded just by going far from the cities, it’s the same sky everywhere on earth with the same satellites flying in it. As it is, all humans live on Earth, so it’s scary to think of how one man like Elon Musk can effect all of humanity.

I doubt Elon Musk will destroy the world, but existential threats aren’t the only thing that matters. More important than living a long life is living a good one, and not being able to see the stars decreases our quality of life. Seriously, have you ever seen them before? Do so while you still have a chance, you can’t any city. You don’t know what you’re missing until you’ve actually seen them. Nick Bostrom’s totalitarian state would have an even more dire effect. Sure, maybe it could save humanity from death, but would anyone actually want to LIVE in such a society? The point of the game One Chance that I mentioned earlier is that you have one chance. This is often interpreted as one chance to save the world from the existential threat, but it’s not. It’s one chance to live your life by taking whatever choices you can, and saving yourself by finding a cure entails neglecting others as every choice has it’s consequences. The same is with life.

At this point it’s probably too late to convince Elon Musk of the error of his ways, but Must is not the only person who has been swayed by Bostrom’s ideas. His ideas should be opposed because they are wrong, and bad ethics has bad consequences. I have never encountered a philosopher as consistently wrong as Nick Bostrom, at least not one as popular who is still alive. As fun as it is to mock him, his influence is genuinely scary. I think the reason Bostrom’s ideas are so popular is because they appear to our most primal fears. He appeals to our fear of death, but goes in the opposite direction of most philosophers. Instead of helping us accept death, he says there is something we can do about it, creating a sense of urgency as nothing else matters, just pursuing this one goal. It’s all snake oil though, as it’s based on logical contradictions and metaphysical claims that are not supported by our understanding of physics or any other science.

If one were to accept nihilism in the face of death then pursuing Bostrom’s philosophy would be as harmless as pursing anything else, but if anything matters then it is important to do what aligns with truth. I maintain we live in a world where neither nothing really matters nor nothing else matters, but things in general matter. The dangers of Nick Bostrom’s ideas are not as catastrophic as the dangers he proposes if you don’t listen to him, but the represent a risk all the same. Maybe we are just drunk ghosts traveling through space, time, and simulations, but even if we are it’s important to consider what really makes life living rather than falling for uninformed fear mongering about existential threats.

TL;DR: We should create a robit that produces paper clips and then sacrifice ourselves to it because such a robot would literally be God and such a sacrifice would be it’s will.

[notabigrock]

This is true though. I remember learning this from a raptor expert at the Cleveland Natural History Museum, not a conservative by any means. The birds who are killed by wind turbines are not the same birds with fly into windows or get covered in oil. This problem is a classic example of the fact liberals (and leftists) are just as bad as conservatives at accepting scientific information that does against their preconceived beliefs, spread here by the Tumblr misinformation mill that appeals to common sense and not sources. The only reason conservatives appear more biased is because their interests are more likely to go against science.

objectively the funniest moral panic to come from conservatives is "alternative energy is bad because birds might fly into the wind turbines". birds, which notoriously have no other man-made obstacles, such as the window

[notabigrock]

Reblog this to ease the back pain of the person you reblogged it from

[notabigrock]

I like how no rush mode is hard mode since you’re not allowed to rush.

[notabigrock]

Sounds like perishable currency was just chosen for flavor, but it has significant economic implications.

First thing to note, cocoa beans aren’t just perishable, they are also consumable. Perishability could be in reference to it’s function as a consumable or as currency, but the prior isn’t really relevant as that’s not how is being used. From what I understand, the Mayan dried their beans so they ceased to be perishable or consumable, so they functioned as regular currency. All four combinations of perishable as a currency and consumable have their own implications though. We already have currency which is not consumable or perishable, so let’s consider the other combinations.

First off, if the currency is perishable but not consumable then it doesn’t really function as a currency. This is because no one wants currency that is going to lose its value soon, and by induction it will have no value ever - not unless there is a significant portion of foolish people, but it’s not something that sustain a currency. However, you could use it instead as a transferable unit of debt with an expiration date. The key difference is such debt needs to ultimately be cashed back to the person who issued it or it’s value will be lost.

If a currency is both perishable and consumable, then the problem is avoided by people consuming the currency before it expires. It’s not a great currency since it’s not a long-term store of value, but it does function as one before it’s expiration date approaches, where it will then function as a commodity instead because people won’t accept it unless they plan on consuming it themselves.

A currency that is consumable but not perishable is the most stable of the tree, but it still has the problem of the money supply decreasing as the currency is consumed. Generally the is balanced by the money supply growing through creation of the currency, but that makes it chaotic in both directions. One of the main reasons precious metals are a common currency is because the amount in circulation is fairly stable. They also have the advantage over cocoa beans as well as other stable assets like diamonds and other precious gems in that they are fungible. In a fantasy setting it’s easy to invent another material with the desired properties, but there is a reason gold dominates, and if using a less desirable currency it may be worth factoring the weaknesses into the plot.

Small fantasy worldbuilding elements you might want to think about:

A currency that isn’t gold-standard/having gold be as valuable as tin

A currency that runs entirely on a perishable resource, like cocoa beans

A clock that isn’t 24-hours

More or less than four seasons/seasons other than the ones we know

Fantastical weather patterns like irregular cloud formations, iridescent rain

Multiple moons/no moon

Planetary rings

A northern lights effect, but near the equator

Roads that aren’t brown or grey/black, like San Juan’s blue bricks

Jewelry beyond precious gems and metals

Marriage signifiers other than wedding bands

The husband taking the wife's name / newlyweds inventing a new surname upon marriage

No concept of virginity or bastardry

More than 2 genders/no concept of gender

Monotheism, but not creationism

Gods that don’t look like people

Domesticated pets that aren’t re-skinned dogs and cats

Some normalized supernatural element that has nothing to do with the plot

Magical communication that isn’t Fantasy Zoom

“Books” that aren’t bound or scrolls

A nonverbal means of communicating, like sign language

A race of people who are obligate carnivores/ vegetarians/ vegans/ pescatarians (not religious, biological imperative)

I’ve done about half of these myself in one WIP or another and a little detail here or there goes a long way in reminding the audience that this isn’t Kansas anymore.

[notabigrock]

Friends, A grassroots movement is calling on all Americans to abstain from shopping with major retailers — including Amazon — tomorrow, February 28, as part of an “economic blackout.” The purpose is to send a clear message: We have the power. We don’t have to accept corporate monopolies. We don’t have to live with corporate money corrupting our politics. We don’t have to accept more tax cuts for billionaires. We don’t have to pay more of our hard-earned cash to Elon Musk or Jeff Bezos or Mark Zuckerberg or the other billionaire oligarchs. We don’t have to reward corporations that have abandoned their DEI policies to align themselves with Trump’s racist, homophobic, misogynistic agenda. We have choices. Most Americans are struggling to keep up. Most live from paycheck to paycheck. Most can barely afford housing costs, food prices, and pharmaceuticals — kept high by monopolies, and fueled by private equity. If politicians won’t hear the voices of average Americans who are being shafted by corporate America, we have to deliver our message to corporate America directly. From midnight tonight to midnight Friday night, please: No Amazon, no Walmart, no Best Buy, no Target, no Disney, no Google, no Facebook. Don’t spend on fast food, major retailers, or gas. Avoid using credit or debit cards to make nonessential purchases. Buy essentials such as medicine, food, and emergency supplies, of course, but make those purchases at small, local businesses. Consider this a test run. If lots of people participate, I’m sure a longer one will be organized. (Tomorrow’s economic blackout is an initiative of The People’s Union USA, which describes itself as a “grassroots movement dedicated to economic resistance, government accountability, and corporate reform.”)' I hope you'll join. What do you think?

[notabigrock]

Jewish people weren’t the first target of the Nazis - disabled people were. Aktion T4 was the first step in the Nazi eugenics program. Everyone knows the Nazis targeted Jews, and indeed they had the highest body count, but people forget that other groups were targeted as well.

[notabigrock]

More!

this is fred, the dot.

fred wants to grow into a beautiful tree, but sadly has no branches

reblog to give fred a branch

i will post fred status updates as he grows

[notabigrock]

Facebook supports end-to-end encryption, where they literally can’t read DMs if it’s activated, but you need to make sure it’s activated. Signal meanwhile is entirely based around end-to-end encryption. IiRC WhatsApp is end-to-end encrypted as well.

PSA: never discuss private affairs in your DMs, especially contraception and abortion. Social media moguls will absolutely sell you out to the government. There are already cases of people being charged based on evidence in their DMs.

[notabigrock]

Posting here because this person thought it would be a good idea to send me a message then immediately block me. That's a good way to get your private conversations made public. For context, this conversation started in the public on the comments of an extremely ableist post that people mistook as inspirational, and the author of the post proved to be a nasty person, insulting people in the comments and blocking people who actually had meaningful dissenting opinions while leaving blatant trolls (eg. "stealing from poor people is good!") be. Due to the ableist context of the conversation that the conversation was born in, I think it's an important one to have in order to fight said ableism. Of course said author blocked me, forcing the conversations to be moved private, so it's only fitting they are made public again by the same process. First off, it's amazing how disconnected from reality these people are. If someone were to somehow capture the image generated by a kaleidoscope and publish it, of course that person would be considered the artist - no sane person would consider the kaleidoscope to be an artist. It would be a type of photography, and photography has been recognized as a form of art for a long time even though with most photographing the artist plays a less active role because they didn’t manipulate the environment like how this hypothetical artist manipulated a kaleidoscope. An “AI Artist” has even greater control, as while the kaleidoscope is used for generating kaleidoscope-type images which are difficult to predict how they are changed with each tap, an “AI Artist” has access to a full semantic space of prompts that directly relate to the output. This is like arguing that Jackson Pollock didn’t make his works because he just threw paint until it ended up like what he wanted to make, so actually the paint made it. Paint is a tool, a kaleidoscope is a tool, a generator is a tool, a person is an artist. The whole argument that the AI made the art instead of the prompter relies on anthropomorphizing the AI, but apparently people have gotten so caught up in the argument without regard for its validity that they are now anthropomorphizing kaleidoscopes.

Now to the argument about the amount of work. This is a completely different argument than debating authorship, and the fact it’s part of the same paragraph leads me to believe they were actually making a third argument fetishizing manual labor (which is so much more ableist) by juxtaposing it with “effort of thought”, but let’s tackle this argument first. I decided to do an experiment - can I draw a picture of a butterfly that better fits my artist vision in less time than a generator? So I prompted Craiyon to make a "butterfly" and after about a minute it made this:

While it was generating though I drew this in Paint, and finished before generating completed:

The butterfly I drew better matched my artist vision, which should be clear from the fact it actually resembles a drawing, and I created in less time if not with less effort. The point should be clear - using a generator to make art does not inherently take less effort. This is not to imply the generated image of a butterfly is art - it is not, but to make the image actually resemble what I want would take more effort and more time. Of course, the idea that more effort automatically makes something more artistic is blatantly absurd. For example, I could have made the same image of a butterfly by editing RGB values in text instead of using Paint, but it would just be a less efficient way to achieve the same result. All that is important is conforming to artistic vision, and wasting time by using inefficient tools should not be fetishized.

As to the idea that physical labor is what makes art, not only is it nasty and ableist (though you can't expect better from someone who thinks disabled artists aren't "real artists"), but it's not a position anyone in the actual art would would take seriously. For example, I am an artist who had a collage of mine professionally exhibited. However, due to my poor fine motor control I technically didn't create the final artifact myself - I assembled all the images I was using into the final image, then my mother tapped it onto folders to hold it together. I could not do this myself without it falling apart. I was completely open about how I made the work with the institution running the exhibition, and they credited with my name solely because I was obviously the artist behind the piece. Fetishizing physical labor is even worse than fetishizing time since at least the latter isn't denying some people access to art since everyone has time (though whether they have free time as a result of economic conditions is a different discussion). To deny disabled people art is to deny their humanity.

The fundamental problem with these people is they've gotten so absorbed into the economic systems that commodify art that they've forgotten what art even is, and that is just sad. This is why they feel the need to insert references to commissions even where the analogy is inappropriate, because that is the lenses they are trained to view art production through since it's the most promising venue for artists *on Tumblr*. Of course people who make their livings through commissions deserve to be able to make a living, but there are whole worlds of art outside of these artists. As one example, it's pretty hard for someone using "AI" to be competitive as an artist at an art show for literal paintings since you'd need an actual robot to actually paint with an AI. While it's certainly true that sometimes people turn to image-generators instead of commissioning art, that's not what I was discussing. I was referring to people who use generators because they want to make art *themselves* since self-expression is a human need, and the text based interface allows them to engage with the creative process in a way they weren't able to before, for example, because having apraxia made them need a discrete interface. I don't think the current wave of text-image generators are ideal for this function, but they do fill it in ways that commissioning an artist does not.

Now, if you're actually concerned about the economic impact of "AI" on artists, then demanding people have to pay for it is the absolute worst thing you can suggest. I get the intention was to strong-arm people into commissioning artists instead of using generators while completely ignoring all the other use cases of generators, but it wouldn't even be effective for that intention. The reality of the situation is the competition isn't between "real artists" and "AI", but between artists who use all the tools available to them and those who refuse to. The key thing to keep in mind is that as long as tools are available, artists are always going to have the competitive advantage at making art over anyone else using the same tools since they can create at least as good artwork in the same amount of time, and better artwork with more time. If nothing else, artists can modify whatever a generator spits out by hand to get it to a much higher level of polish. However, this inherit advantage would be lost if artists lose access to tools, since people who can afford the tools would then be able to make things faster than those who can't, giving people with money a competitive advantage over artists so unless someone really needs the higher quality output they will pay to use the generator instead every time in order to save time - or in this case, a grifter with access to "AI" just enough art skills to outcompete actual artists who can’t afford to use “AI” since they can do the same job in half the time. Even if the price is jacked up so high that it's more expensive than commissioning artists, the time advantage is enough that big corporations will still pay to hire those who use AI over those who don't. I know artists who have worked at Disney and Epic for decades, and the message I got from them is clear - "AI" isn't going to take your job, but someone who knows how to use "AI" will. It's a sad truth, but the fact of the matter is these jobs always required learning the latest tools to stay competitive, and "AI" is just the latest in the long line of tools, with it only being the greatest disruption since CGI.

I understand the frustration artists in this community have, but pushing for policy to spite disabled people who wouldn't even be paying for commissions anyway is not the solution, especially when such solutions would also hurt artists who do commissions. For more thoughts on "AI Art", including actual solutions to practical problems, I have a full blog post here:

PS: "People like you just just want to abuse it. " Bruh, I don't even make "AI Art", I'm just calling out blatant ableism where I see it. But hey, go ahead and prove you're a nasty, spiteful, judgmental person who utterly lacks reading comprehension and is quick to make assumptions with no evidence.

[notabigrock]

The audacity to claim that your life was ruined when you just kicked someone out of his house, stole all of his money, and turned his parents against him.

Oh come on lady, you can't deny a man his gaycation

[notabigrock]

First things first, despite the name of the blog, I am in fact @abigrock - the not is referring to "not that blog" rather than "not that person".

As someone who has been on this site for over 13 years and is a computer scientist whose Master's thesis was on recommendation systems, one thing I'm going to emphasize is you CANNOT trust the community. Aside from outright lying, misinformation spreads extremely quickly due to people reblogging sensationalist content without factchecking. You can't even trust them to understand the functioning of the site, since how popular bloggers use the site is very different from how the site is designed. For example, you may have noticed people putting extended messages in the tags. This is as a result of an unintended affordance of tags - the fact tags aren't included with reblogs means people use them to send messages they don't want carried over in reblog chains instead of using them as an actual tags. As result of this abuse of the feature noise is added to the recommendation system Tumblr is based on which is centered around tags, but it doesn't cause so much damage that the recommendation system ceases to work as intended. Indeed, a new user to the site will be subjected to recommended content based on information from popular tags immediately.

I just conducted an experiment where I created an entirely new account to see what the user experience is like and spoilers, it's not anything like what the community describes. Before you even create an account, you'll be subjected to a feed of trending posts. I'm willing to bet this content is based primarily on likes, not reblogs, since likes give a clearer signal that the intention of the interaction was expressing interest in the post, when reblogging can be done for countless reasons, including engaging to express disapproval. Once the an account is created, the first thing that is asked from a user is to follow 5 tags, with trending tags and that top, followed by tags in different clusters arranged alphabetically. Based on the tags. the user is then asked to follow three blogs. After this, the user is brought to the explore feed. ALL of these steps are recommendations based on machine learning, but the last step is very explicitly so. The site couldn't be less subtle after I liked my first post:

Note that contrary to what was claimed, the algorithm is not slow to learn. This recommendation was given after only the single previous like, where previous I had shown no interest in food based on my followed tags or blogs.

It's also important to note that switching to from "explore" to "home" and then from "for you" to "following" does not cause the recommendations to go away. While the first few posts in the feed are from followed blogs (though not in chronological order as some have claimed - can't prove it with screen shots since the site is designed to only show one post at a time, but it's easy enough to check just by scrolling for a bit. it doesn't take long to see posts out of order), you'll start seeing recommendations before long.

As far as I can tell, people are conflating the communities' emphasis on reblogging (because encouraging reblogging is the most effective way for a creator to increase their reach) and use of features to support that engagement style with the actual functioning of the site. If the site did not value liking posts for the purpose of informing it's recommendation system, it would not even have a like feature in the first place - this is clear from how clunky it is to navigate likes as compared to navigating blogs. One critical thing to be aware of is that if you use the site in the way the Tumblr community prescribes *you're not going to see all the other ways countless people use the site*. I'm willing to guess the majority of users, especially new ones, aren't even aware of the "Tumblr community" since their posts are not the ones being pushed by the algorithm, and interacting with it requires delving through notes instead of just passively consuming content. Tumblr is much more like other sites than people realize, the "Tumblr Community" just doesn't realize that because the people who use the site in that manner are not the people interacting with the "Tomblr Community".

PS: As a computer scientist I'm really curious as to what laymen think "algorithm" means since whatever people are describing as algorithms are not what I'd consider algorithms since whatever they are they lack key properties of algorithms. I especially wonder why people incorrectly use such a long and clunky technical term when I'm sure there are much easier, common words that more accurately describe whatever is being referred to.

"lol tiktok users would rather learn mandarin than come to tumblr" yeah man that's crazy. who would have thought that people from the video-based dopamine-machine app wouldn't decide to join the text- and image- based no-algorithm website?

[notabigrock]

Missing like, actual Christmas songs about Jesus overcoming death which are metal AF.

an incomplete taxonomy of christmas music

[notabigrock]

I don’t know of any congregations that actually have teams, it’s just that people in the congregation get together to play sometimes, usually Tuesday nights.

Me, Catholic, walking into a Protestant church with no depictions of Mary: where's my mom

[notabigrock]

My favorite was seeing them try to play off a black rat snake as a black adder for My Lady Jane.

it is SO FUNNY when you’re watching a show with horses in it and within the plot of the show the horses are “”””upset””” but the way they communicate this is by placing a nervous nickering sound bite over what is clearly a calm horse placidly performing a practiced trick

[notabigrock]

I’ve always taken sloth to mean willful dereliction of one’s duties. It’s definitely not the same thing as laziness, which is an aversion to action and is just a personality trait, not a sin.

Hey who else here thinks sloth is a dumb sin and just sounds like mental illness raise your hands

[notabigrock]

Parsing tone is not the issue, it can handle tone as well as any other aspect of text since it’s all there in it’s training data. The actual issue is that it has no concept of truth. All it does is generate text based on existing patterns found in text, it’s not trying to achieve any objective beyond that. The reason I find it important to counter this with the actual reason for its incompetence is the incorrect explanation is based on myths conflating autistic people with robots, and is then in turn being used to argue for incompetence of autistic people in executive positions. Lots of people have trouble parsing tone, especially online, but that’s fine as long as they have a basic understanding of reality and intend to communicate information that is true and helpful, which does not apply to these AI.

i’m fucking dying. it’s been brought to ben collins (the rad new owner of the onion’s) attention that Google AI summary is using The Onion headlines, resulting in this: