One of the occasional defenses of generative AI is that it quote-unquote ‘democratizes’ art and writing — and then, as with the NaNoWriMo statement yesterday, it becomes somehow problematic to condemn generative AI, because what, do you hate DEMOCRACY? Do you not want everyone to have access to art and writing? Oh! Oh! Somebody doesn’t want the competition, doesn’t want the masses to rise up with the FREEDOM of their RENEWED ACCESS to ART and STORY, you PRIVILEGED ELITE BASTARD.
But I think it’s important to take the air out of these things (often by kicking the absolute shit out of them).
Generative AI is not democracy.
Generative AI is not free.
Because that’s the cornerstone of the idea, right? It’s a freely accessible tool that evens the playing field.
But generative AI has considerable costs.
Let’s go through them.
1. Money, Cash, Ducats, Coin
Access to much of generative AI will cost you actual money in many cases, though certainly it’s also becoming freely accessible at some levels — and more and more services are forcibly cramming it into their existing platforms, which, I’d like to note, is seriously fucking annoying. I’m waiting for the day where my microwave tries to write and sell its “slam poetry.”
Still, free now isn’t free forever. I mean, the “first taste is free” drug deal rule applies here, c’mon. They get you interested, you use it, and suddenly it costs more, and more, and then more again. They have to do this. The development of this fucking nonsense horseshit has been a billions-of-dollars investment. They want that money back, and if that means they have to put it on a chip and have Elon Musk fire it into your skull with a modified .22 rifle, then that’s how they’ll do it. If it remains free to use, then that means it’ll come with advertising jackhammered into it. (“Every time I ask it a question, it answers ‘Taco Bell Crunchwrap Supreme,’ wtaf.”)
2. Future Money
Generative AI is meant as a disruptor. And classically, disruption is not always a good thing. (One might argue it’s rarely a good thing.) Big shiny new tech company shows up, reinvents a thing by offering it cheaply and loopholing its way around regulations, you get hooked, the older industry withers on the vine, the shiny new tech company nests inside the chest cavity of the older industry until its dead and it can erupt out from the carcass in a spray of blood and bone, and then it just charges you even more than the older industry did for what may potentially be a lesser product.
As such, the way one can currently earn money from art and writing is at risk thanks to the rise of generative AI. How this might happen is myriad — Amazon getting flooded with AI books makes it harder to find any book; companies learn they can generate “content” with the push of a button and either choose to do so or use the threat of doing so as leverage to reduce the money they will pay for art and for writing; generative AI’s implementation damages enough outlets for art and writing and sends them packing, which means fewer outlets for artists and writers, which lowers opportunity and, by proxy, money; generative AI acts as a labor scab during union disputes for creators; writers and artists are no longer hired to iterate and create but rather to “edit” and “fix” the work “created” by generative AI, which is to say, generative AI artbarf robots puke up a bunch of barely digested material and a company pays a cut-rate to once-notable writers and artists to push that slurry into some kind of shape, like they’re Richard Dreyfuss with the mashed potatoes in Close Encounters of the Third Kind.
And that’s just a sampling.
Ultimately, it puts power in the hands of corporations and tech-bros, and removes the power from artists and writers. And will try to eat away at copyright laws to do so.
That’s not democracy. And it certainly doesn’t come free.
3. Future Artists, Future Writers
There is a literal human cost. There will be people going forward — and, I’m betting, there are people right now — who are going to turn away from the art-and-writing path because of this. I know kids who already look at those career paths with the question of, “What’s even the point?” There will be a bonafide brain drain from the bank of artists and writers. (Not to mention teachers, or any other career currently being targeted and poached by generative AI on behalf of awful corporations.)
(And here, my conspiratorial eye-twitch red-thread-on-a-bulletin-board personality comes out and says, well, that’s awfully convenient — we’ve already seen such a heavy lean into STEM and away from the Humanities, because artists and writers tend to be thinkers, philosophers, they tend to have empathy, they tend to be less interested in the hustle culture churn of corporate life, and this only drives that nail in deeper, doesn’t it?)
Again, doesn’t sound like it’s democritizing shit. Anything that makes it harder and less likely to become a thing isn’t democritizing that thing.
4. The Costs of Actual Theft
Uh yeah, it steals shit. That’s how it works. It can’t do it without stealing shit. They’ve admitted it. Out loud. I don’t know how to explain to you the very real cost of having your work yanked out of the ether and thrown into the threshing maw of generative AI so your creations can become hunks of fake meat in their artbarf stew. But the cost isn’t metaphorical. It’s literal.
Once again, that’s not democritizing anything. It’d be like saying, “Ahh, Google has stolen your vote, and will vote on your behalf. How wonderful! You don’t even need to do it, now. We’ll handle it for you, for free. See? We’ve democritized democracy!”
God, even as I typed that out it feels alarmingly possible.
*shudder*
5. Environmental Cost
You don’t need to look far to learn about the environmental costs of generative AI. We didn’t ask for it, but it’s here, and even casual use can increase the burden on our environment.
We’re in danger of turning away from our already too lax environmental goals. We need coal and other fossil fuels gone, we need to protect water usage, and here comes AI to gobble up the water and our power and force us onto our back heel, all because some dickheads want a robot to lie to them about how many giraffes they see in Starry Night or because they need the magic computer to draw for them a picture of a 13-fingered Donald Trump freeing White Jesus from the cross with a couple of M-16s.
The only thing that’s democritizing is the death of our natural environment. Wow, nice work, Tech Bros. Guess that’s why Google removed their plan to DO NO EVIL from their mission statement.
6. The Damage to Informational Fidelity
It is increasingly hard to tell truth from fiction. Visually, textually, it’s getting easier and easier to just… lie, and to do so with effective facsimiles made from generative AI. Trump posting that Taylor Swift endorsed him, or creepy videos from Twitter’s AI showing Kamala Harris covered in blood and taking hostages, so newer abilities on a phone to just take an image and edit in whatever you want with the touch of a button — a giraffe, a bloody hammer, a hypodermic needle, a child’s toy, a sex toy, a loaded gun, whatever. The laws are far far too slow to catch this. This will be propaganda, given a nuclear-grade steroid injection. This will be revenge porn, god-tier level.
To sum up?
AI isn’t free.
It isn’t sustainable.
It isn’t democratizing a damn thing.
The tools and skills to create are already available. No, not perfectly, and no, the industries surrounding art and storytelling are certainly imperfect. But AI doesn’t push the existing imbalance into the favor of artists and writers, but rather, the opposite. And as it does so, it burns the world and fucks with our ability to tell truth from fiction, even right from wrong.
It’s weird. It’s horrible. I kinda hate it. I hope we all realize how absolutely shitty it is, and we can eventually shove its head in the toilet, same as we did with NFTs and crypto. Shove it in, give a good couple flushes.
In a Telegram channel with around 1,300 members, there are individual chat rooms for 70 colleges and universities across the country. The bolded text are names of universities. (captures from Telegram)
“Does anybody know *** ***-***, born in 2007, from *** High School? Send me a DM if you do.”
After recent revelations about sexually explicit deepfake images being shared on certain Korean university campuses, the Hankyoreh has uncovered evidence that the illegal practice of creating manipulated images of acquaintances from school or a particular neighborhood and distributing them is occurring rampantly across the web. Those targeted by the deepfakes include students not only at the country’s biggest colleges and universities, but students and minors at middle and high schools as well.
The Hankyoreh conducted a web search on Aug. 21 scouring various social media platforms, revealing numerous Telegram chat rooms that distribute illegal deepfake pornography categorized according to region of residence and school. The crimes are continuing to expand in breadth while becoming more detailed.
The process is as follows. A group funnels potential members through a Telegram chat called “friends of friends.” From there, groups are further divided into regions of residence and university. Members then chat with each other about specific girls and women to see if multiple members recognize any of them. They then acquire ordinary photos of their victim from social media and utilize them to illegally produce pornographic deepfakes.
One Telegram channel has over 1,300 members. This channel hosts various chat rooms that are categorized according to university, with over 70 universities represented. Members post photos of young women they know, along with basic information such as their majors, when they started school, and their names. Other members within the same chat room then chime in if they know the young woman. People who know the person then form their own chat room to produce and share illegal content. Once there are enough deepfakes for a specific person, members will create chat rooms dedicated solely to that individual. The chat rooms are labeled with names like “Degrading Kim ***-***.”
The deepfakes are distributed systematically on a disturbing scale. The sexually explicit illegal images are shared in a chat room over and over, where some members even create emojis out of them “for fun.” The images created in one room quickly spread to other chat rooms.
One chat room, called “Link Sharing Space,” had over 3,700 members. Members of this chat room share links leading to dedicated chats for deepfakes specific to one person or deepfakes pertaining to students from a specific university. Some chat rooms even have selection processes, where applicants must submit 10 photos of someone they know and pass an interview before being allowed in.
One Telegram channel, dedicated to deepfakes of middle and high-school students, had over 2,340 members. Members of this channel regularly produce and distribute content that lands you at least a year in prison for simply possessing or viewing it.
The criminals who produce such disturbing deepfakes are widening the breadth of their victims while becoming more detailed in the classification and labeling of their illegal content, adding to mounting fears shared by many women and girls.
“Although I’ve made my account private, I can’t stop thinking about someone I know using my photos for criminal purposes,” a woman, 24, who normally enjoys using social media platforms told the Hankyoreh.
Another 24-year-old university student, who asked the Hankyoreh to only identify her by her surname, Kim, said, “It’s impossible to know where the photos are being shared, and how far they’ve spread. It’s truly depressing.”
Since Telegram is an overseas firm with servers outside South Korea, the Korea Communications Standards Commission and other domestic agencies do not have the legal jurisdiction to demand the erasure of content being distributed through its chat rooms. Even if local authorities conduct an investigation, the confiscation and or search-and-seizure warrants they attain have no teeth, making it extremely difficult to identify individual suspects.
“Deepfakes distributed on Telegram can then spread to other websites, where they are reproduced or altered and redistributed. We therefore need to revise laws like the Telecommunications Business Act to require any overseas platforms with branches in South Korea to cooperate with domestic investigations,” said Min Go-eun, an attorney.
Critics also point to the lukewarm attitude of investigative authorities and the ineffectiveness of legislation.
A Telegram channel for “middle and high school friends of friends” with 2,340 members. Once a user verifies their identity, they find common acquaintances and generate deepfakes. (capture from Telegram)
“The most common complaint we hear from victims of deepfakes is that investigators cannot identify specific suspects operating via Telegram, but private actors such as ‘56 Flame’ have identified specific individuals in connection with deepfake crimes,” said Heo Min-sook, who researches legislation for the National Assembly Research Service.
“Investigators are essentially twiddling their thumbs and not exploring all possible routes of identification,” Heo added.
“Someone who produces illegal deep fakes with intent to distribute can be sentenced to up to five years in prison or a fine of 50 million won (US$37,453), but the majority of suspects are given probation, or their indictments are delayed,” Heo continued.
“We need to revise legislation to punish not only those who distribute or produce such deepfakes but those who possess or view it as well. The courts also need to realize the severity of such crimes,” she said.
In the wake of Biden’s withdrawal from the race and Kamala Harris’s shockingly rapid and decisive ascension, I have begun feeling emotions that I haven’t allowed myself to feel in connection to politics in a long time: hope, excitement, even enjoyment. As many commentators have noted, there is a joyfulness, even a level of fun, to Harris’s campaign that is an almost shocking contrast to what came before. The fact that they are playing along with the JD Vance couch meme may be the clearest sign that they are in tune with contemporary culture, but the more general pattern of calling conservative leaders “weird” and “creepy” feels like a major turning point — not just in terms of political tactics, but in terms of liberal political culture. It marks the end of a certain fatalistic defensiveness on the one hand, and also of the joylessly self-righteous habits of policing and shaming allies on the left.
For all my life, conservatives have been the norm. Everybody (who matters) feels at best very uncomfortable about abortion and non-normative sexuality. Everybody (who matters) resents the burden of funding high-quality public service. Etc., etc., etc. Tactical observations from the early 90s hardened into inescapable truisms, even as they became less and less true. This produced a permanent defensive crouch, as Democrats seemed to believe that Republican rule was the norm and they could at best eke out a narrow win to take their turn — at passing a more nuanced and “smarter” version of Republican policies. Priority number one after each victory was to get bipartisan support, as though Democrats didn’t believe it was legitimate for them to legislate on their own. Attachment to the fillibuster rule among the older cohort of Democratic senators is the most destructive example of this built-in defeatism.
So the new confidence of the Harris campaign is refreshing, as is the contempt and puzzlement they express at conservatives’ increasingly unpopular and downright bizarre beliefs. More specifically, what is refreshing here is their willingness to be mean, to insult, to reduce their opponents to sputtering speechlessness. I cannot emphasize enough how much of a break this is with the joylessly self-righteous policing of language that has been the norm among liberals and leftists for my entire adult life. From that perspective, the parody responses write themselves — “we shouldn’t kick-shame JD Vance…” or “Republican leaders won’t read your post calling them weird, but your friends who could be viewed as weird for completely unrelated and totally harmless reasons will…” — and the fact that it’s so easy to come up with them shows how that mode of engagement has reduced itself to self-parody.
A key shift for me when I saw a white man worrying aloud about the tendency to refer to Vice President Harris as “Kamala” — there is of course a whole history of belittling people by refusing the respect of their last name, it’s especially fraught since she’s a Black woman, etc., etc. And I will be honest with you and say my first response was that this person should simply shut the fuck up. “Kamala” is a very distinctive name, whereas “Harris” is not. Her own social media team is called “Kamala HQ.” We do not need to get out ahead of the supposed “victim” herself.
More broadly, though, my strong gut reaction reflected my belief that we have just got to be done with this style of whiny preemptive strike against any hypothetical offense that may one day be perceived. Political correctness is a strategy that has failed. Aside from eliminating the grossest slurs and overtly bigoted jokes — which even conservatives themselves know not to share in mixed company — it has produced only irritation and insecurity.
People like to present it as simple common sense, but the euphemism treadmill and, more than that, the constant incentive to find ever more nuances of linguistic “oppression” ensure that the politically correct linguistic norms can never actually settle into a coherent common sense. It produces bristly defensiveness in those who can’t keep up and an unhealthy and counterproductive readiness to be offended among the avant-garde. My favorite example of the counterproductive nature of such language policing is the fact that disability activists are more or less singlehandedly keeping alive the etymological association of words like “moron” with disability. Linguistic usage moves on — take the win! But no, etymology is destiny when it gives you something to nitpick and alienate potential allies over.
Related here, I think, is the culture of constantly nitpicking headlines from the New York Times on social media. Again, the belief is that politics will take care of itself if everyone agrees to speak in just the right way. Obviously, the New York Times is a bad actor in many ways and the media establishment is artificially propping up Trump through the application of double standards. They are worthy of critique, but the obsessiveness and detail-orientation of the critique is what raises my hackles. It bespeaks a whiny wounded entitlement, as though we all believe that the New York Times should be a liberal actor or that a simple description of reality would automatically favor our politics.
In reality, to do politics, you have to do politics. The media is not an umpire, it is a terrain of struggle. The Harris campaign is engaging that struggle much more effectively than Joe Biden ever could because it does not embrace the false premise that the New York Times is or should be on her side by default. And she has received overwhelmingly favorable coverage! Not 100% — there is always something to whine about, of course. They don’t always put the word “falsely” in the headline, and sometimes they take too long describing Trump or Vance’s claims before debunking them. They achieved that not by whining about how unfair the Times has been, but by actively setting the agenda and setting the tone.
And that tone is mean. It is contemptuous. It aims to harm Trump and Vance and their reputations. It aims to make them personally angry and make people question their loyalty to such deeply flawed men. To achieve these goals, it is not overly concerned about petty details like whether JD Vance really engaged in an elaborate form of masturbation involving his furniture and described it at length in his memoir — much less whether other furniture-masturbation enthusiasts might be collateral damage of the joke. The joy and humor and fun of the Harris campaign, the way that it acknowledges the Republicans as enemies without setting them up as all-powerful, hopefully marks a decisive end to that kind of idiocy. From now on, entitlement and prickly defensiveness can remain in its more natural home — among the washed up losers who have coasted on white male privilege so long and spent so much time in the “safe space” of their ideological bubble that they don’t realize how pathetic they appear to anyone halfway normal — because eventually everyone complaining about unfair media coverage and moaning about how they aren’t being shown the proper respect will realize that… they sound like Trump.
When starting a new television show, one of two impulses typically dominates. There's "I want to see something comforting." And then there's that reaction's darker, more danger-filled, and beautiful sibling: "I want to see some shit I've never seen." Scavengers Reign falls, quite gloriously, in the latter camp.Originally pitched to Adult Swim, given a shaded existence in the smoke of the rolling garbage fire over at Max, and now getting a bit of genuine sunlight thanks to its move to Netflix and much-deserved Emmy nod, the sci-fi series operates at multiple levels of frequently horrifying joy. But the most basic one is this: Sit down to watch one of its 12 extant episodes, and you're practically guaranteed to get at least one addition to your personal list of the most fucked-up things you've ever seen.Drawn in soft, gorgeous pastels that deliberately evoke the otherworldly landscapes of French sci-fi legend Moebius, Reign concerns itself with the crew of the Demeter, a space-cargo ship that has the ill fortune to suffer a catastrophic failure in the vicinity of the life-rich world of Vesta. And the thing about the Vestan flora and fauna—as the few survivors who make it to the ship's escape pods (while most of their comrades remain locked, semi-safe, in cryo sleep) quickly find out—is that they aren't unwelcoming to human life. Quite the opposite: The Vestan creatures are desperately eager to interface with humanity (as if they need us), whether that means feeding on us, growing inside us, or otherwise adapting us into their all-powerful biological compulsions and needs. You're going to be part of the system one way or another, Vesta asserts, with snapping teeth, subtle poisons, and slowly creeping tentacles. Adapt or be destroyed.It's the show's greatest selling point that it never makes this overwhelming obsession with adaptation, dependence, and the ugly/glorious things that happen at their intersection either academic or dry to watch. Vesta is, in point of fact, a decidedly wet world, whether you're watching a psychic scavenger force feed black, oily goo into the mouth of a human survivor, or the show's most adaptable pair, Sam and Ursula (Bob Stephenson and Sunita Mani), climb through a feeding tube/ovipositor to shelter from a lethal storm in the…womb?…of a sea-faring jellyfish-turtle-thing. (Descriptors get tricky when applied to as varied and fascinating an ecosystem as Vesta sports.) And that's before we get into the various outright predators and parasites who appear across the planet's landscapes, which periodically inject a bit of shockingly bright red into the show's otherwise soft palette with sudden gushes of arterial spray.We found ourselves wondering—while re-watching the show's single (and at present final) season on Netflix this week—if James Cameron has seen this shit. Scavengers Reign often feels like a reaction to Cameron's work. Its opening scenes, and the recurring reminder that "The Company" couldn't give two shits about its wayward employees, underline the bureaucratic and capitalistic indifference that gives such awful weight to the Alien franchise's blue-collar vision of space. Vesta, meanwhile, feels like a far more creative, visceral, and real version of Cameron's too-friendly paradise of Pandora. There is no easy "Our hair connects to their tails!" melding of different biologies here. Even the most beneficial symbiosis in this world comes with undercurrents of horror, the creation of something new requiring the loss of something old.None of which would work, we'd argue, if the show neglected its human element. It is to the credit of the series' writers, and especially its cast (which also includes Wunmi Mosaku, Alia Shawkat, and Ted Travelstead) that the survivors, while a bit thinly drawn, come across as real people in a terrifying, occasionally wondrous situation. One of the show's deeper aims is to demonstrate the way human beings already exist tangled up in systems: co-dependent relationships, guilt complexes, the larger social and financial structures that prey on us each in turn. (When the show introduces a new trio of characters fairly late into its run, it does such a good job of laying the groundwork that it's impossible not to see them as another form of invasive life, complete with its own unwritten laws and ruling hungers.) And without liking characters like hardened survivor Azi, or even an increasingly deranged martyr like Travelstead's Kamen, none of the beauty or the horror of those various biological or mental traps would sink its teeth into the show's viewership.This will, Eywa willing, continue to grow. Whether you think Scavengers Reign itself deserves or requires a second season—it ends on a note that could easily inspire one but also closes so many of its circles in a satisfying-enough manner as to not require it—it's something that deserves to be watched for what it already is. Which is like nothing else on TV at the moment: horrifying, warm, philosophical without being pretentious, occasionally funny, and above all creative. Its artists, writers, and performers have put an enormous effort into showing you things you've never seen before. Even when its animation looks inexpensive, it never once looks cheap. It has a new idea lurking around every corner, waiting to snare and fascinate you. If your brain is of the type to be fertile soil for it to plant its seeds in, it will do so aggressively and with gusto. We could potentially live without a second season, but we need this team to have more room to make television shows like this one, one way or another. Comfort is cheap, easy, safe. Scavengers Reign is the other thing.
The [CrowdStrike] catastrophe is yet another reminder of how brittle global internet infrastructure is. […] This brittleness is a result of market incentives. In enterprise computing — as opposed to personal computing — a company that provides computing infrastructure to enterprise networks is incentivized to be as integral as possible, to have as deep access into their customers’ networks as possible, and to run as leanly as possible.
Redundancies are unprofitable. Being slow and careful is unprofitable. Being less embedded in and less essential and having less access to the customers’ networks and machines is unprofitable — at least in the short term, by which these companies are measured. This is true for companies like CrowdStrike. It’s also true for CrowdStrike’s customers, who also didn’t have resilience, redundancy, or backup systems in place for failures such as this because they are also an expense that affects short-term profitability.
A reminder that presentism is perhaps an even bigger threat to our economic infrastructure than it is to our common culture.
Everything I read about Silicon Valley’s support of Trump comes down to this:
These are people who believe in spending massive amounts of public money to enrich themselves while they make shitty tunnels under Las Vegas or drop off scooters on city sidewalks or chase self-driving cars, so firmly believing that these are total solutions ready to solve all problems right now, in complete ignorance of any existing systems or mechanism or solutions that exist.
These are the folks who reinvent busses or trains, but do so in a way that will make them rich, and therefore, they’re better. Except in practice, each time we follow their lead we end up with something worse than what the rest of the world gets through competent government. These are the folks who think the solutions are $600 home test devices for COVID and not wearing a mask. These are the folks who will block real solutions while they waste money failing over and over to solve big problems and then walk away without a consequence. These are the folks who think the only thing we have to learn from each other or other countries is what cannot yet be exploited for profit by a Stanford drop out building something 1/8th as good for 10x the price.
These are the people who think the best things that have happened over the last fifteen years have come out of Silicon Valley, even though virtually all of those things are not profitable and have come with major downsides.
I work in tech. I think a lot of cool stuff is being built and a lot of good work is being done. But tech is a mature industry, and most of what is interesting these days has to do with bringing the things we learned from 2000-2015 about how to use software into places that have not yet modernized. We’re at the tail end of what’s interesting and good and novel. Software technology has very little left to change in a major way. And the entire ethos of a16z and the like has utterly failed to produce breakthroughs in computer hardware, biological sciences, energy, environment or any other major sector. The last decade of innovation has been entirely about reducing friction in commerce. That’s it. And it’s not that profitable and will end up with a very small number of winners.
The major successes in tech are largely SaaS companies selling tooling to hopeful SaaS companies. It’s a spiral-jerk that ends in an easier buying experience online or shitty advertising.
The problems we face in the US, and the problems faced by folks throughout the world, will not be solved on Sandhill Road. And the thing is, they all know this. Support for a monstrous fascist like Trump is the warning sign. It’s just like how companies don’t move to Texas to be great, but instead to squeeze margin out of cost cutting everywhere you can when you no longer capable of growth or innovation. The Trump-Vance ticket has the support of Silicon Valley because their goal is to have government give up. Elon Musk pushed the hyperloop to stop California high speed rail. And in that space, Silicon Valley can try and convince us to drive self-driving electric cars underground. When that doesn’t work, they walk away, and the problem remains unsolved. In the meantime, we’ve wasted billions and they’ve made millions off of carry fees. When the government isn’t even trying, it creates space for charlatans to step in.
Think of all the problems Silicon Valley won’t solve, but can look great telling LPs that they’re part of the solution. Doesn’t it feel better to be part of the solution and make a profit instead of paying taxes? Never mind nothing will be solved.