This weekly roundup thread is intended for all culture war posts. 'Culture war' is vaguely defined, but it basically means controversial issues that fall along set tribal lines. Arguments over culture war issues generate a lot of heat and little light, and few deeply entrenched people ever change their minds. This thread is for voicing opinions and analyzing the state of the discussion while trying to optimize for light over heat.
Optimistically, we think that engaging with people you disagree with is worth your time, and so is being nice! Pessimistically, there are many dynamics that can lead discussions on Culture War topics to become unproductive. There's a human tendency to divide along tribal lines, praising your ingroup and vilifying your outgroup - and if you think you find it easy to criticize your ingroup, then it may be that your outgroup is not who you think it is. Extremists with opposing positions can feed off each other, highlighting each other's worst points to justify their own angry rhetoric, which becomes in turn a new example of bad behavior for the other side to highlight.
We would like to avoid these negative dynamics. Accordingly, we ask that you do not use this thread for waging the Culture War. Examples of waging the Culture War:
-
Shaming.
-
Attempting to 'build consensus' or enforce ideological conformity.
-
Making sweeping generalizations to vilify a group you dislike.
-
Recruiting for a cause.
-
Posting links that could be summarized as 'Boo outgroup!' Basically, if your content is 'Can you believe what Those People did this week?' then you should either refrain from posting, or do some very patient work to contextualize and/or steel-man the relevant viewpoint.
In general, you should argue to understand, not to win. This thread is not territory to be claimed by one group or another; indeed, the aim is to have many different viewpoints represented here. Thus, we also ask that you follow some guidelines:
-
Speak plainly. Avoid sarcasm and mockery. When disagreeing with someone, state your objections explicitly.
-
Be as precise and charitable as you can. Don't paraphrase unflatteringly.
-
Don't imply that someone said something they did not say, even if you think it follows from what they said.
-
Write like everyone is reading and you want them to be included in the discussion.
On an ad hoc basis, the mods will try to compile a list of the best posts/comments from the previous week, posted in Quality Contribution threads and archived at /r/TheThread. You may nominate a comment for this list by clicking on 'report' at the bottom of the post and typing 'Actually a quality contribution' as the report reason.
Jump in the discussion.
No email address required.
Notes -
Award-Winning AIs
To be fair, "best isekai light novel" is somewhere between 'overly narrow superlative' and 'damning with faint praise', and it's not clear exactly where how predominately AI-generated the writing is or what procedure the human involved used. My own experience has suggested that extant LLMs don't scale well to full short stories without constant direction every 600-1k words, but that is still a lot faster than writing outright, and there are plausible meta-prompt approaches that people have used with some success for coherence, if not necessarily for quality.
Well, that's just the slop-optimizing machine winning in a slop competition.
It's a slightly higher standard than isekai (or country music), and Spotify is a much broader survey mechanism than Random Anime House, and a little easier to check for native English speakers. My tastes in music are...
badunusual, but the aigen seems... fine? Not amazing, by any means, and some artifacts, but neither does it seem certain that the billboard number is just bot activity.Well, that's not the professional use!
It's... hard to tell how much of this is an embarrassing truth specific to Studio Larian, or if it's just the first time someone said it out loud (and Larian did later claim to roll back some of it). Clair Obscur had a prestigious award revoked after the game turned out to have a handful of temporary assets that were AIgen left in a before-release-patch build. ARC Raiders uses a text-to-speech voice cloning tool for adaptive voice lines. But a studio known for its rich atmospheric character and setting art doing a thing is still a data point.
(and pointedly anti-AI artists have gotten to struggle with it and said they'd draw the line here or there. We'll see if that lasts.)
And that seems like just the start?
It's easy to train a LORA to insert your character or characters into parts of a scene, to draw a layout and consider how light would work, or to munge composition until it points characters the right way. StableDiffusion's initial release came with a bunch of oft-ignored helpers for classically extremely tedious problems like making a texture support seamless tiling. Diffusion-based upscaling would be hard to detect even with access to raw injest files. And, of course, DLSS is increasingly standard for AAA and even A-sized games, and it's gotten good enough that people are complaining that it's good. At the more experimental side, tools like TRELLIS and Hunyuan3D are now able to turn an image (or more reasonable, set of images) into a 3d model, and there's a small industry of specialized auto-rigging tools that theoretically could bring a set of images into a fully-featured video game character.
I don't know Blender enough to judge the outputs (except to say TRELLIS tends to give really holey models). A domain expert like @FCfromSSC might be able to give more light on this topic than I can.
Well, that's not the expert use!
That's a pretty standard git comment, these days, excepting the bit where anyone actually uses and potentially even pays for Antigravity. What's noteworthy is the user tag:
Assuming Torvalds hasn't been paid to advertise, that's a bit of a feather in the cap for AI codegen. The man is notoriously picky about code quality, even for small personal projects, and from a quick read-through (as an admitted python-anti-fan) that seems present here. That's a long way from being useful in a 'real' codebase, nor augmenting his skills in an area he knows well, nor duplicating his skills without his presence, but if you asked me whether I'd prefer to be recognized by a Japanese light novel award, Spotify's Top 50, or Linus Torvalds, I know which one I'd take.
My guesses for how quickly this stuff will progress haven't done great, but anyone got an over:under until a predominately-AI human-review-only commit makes it into the Linux kernel?
Well, that's just trivial stuff!
I don't understand these questions. I don't understand the extent that I don't understand these questions. I'm guessing that some of the publicity is overstated, but I may not be able to evaluate even that. By their own assessment, the advocates of AI-solving Erdős problems people admit:
So it may not even matter. There are a number of red circles, representing failures, and even some green circles of 'success' come with the caveat that the problem was already-solved or even already-solved in a suspiciously similar manner.
Still a lot
smarter aboutbetter at it than I am.Okay, that's the culture. Where's the war?
TEGAKI is a small Japanese art upload site, recently opened to (and then immediately overwhelmed by) widespread applause. Its main offerings are pretty clear:
That's a reasonable and useful service, and if they can manage to pull if off at scale - admittedly a difficult task they don't seem to be solving very well given the current 'maintenance' has a completion estimate of gfl - I could see it taking off. If it doesn't, it describes probably the only plausible (if still imperfect) approach to distinguish AI and human artwork, as AI models are increasingly breaking through limits that gave them their obvious 'tells', and workflows like ControlNet or long inpainting work have made once-unimaginably-complex descriptions now readily available.
That's not the punchline. This is the punchline:
@Porean asked "To which tribe shall the gift of AI fall?" and that was an interesting question a whole (/checks notes/) three years ago. Today, the answer is a bit of a 'mu': the different tribes might rally around flags of "AI" and "anti-AI", but that's not actually going to tell you whether they're using it, nevermind if those uses are beneficial.
In September 2014, XKCD proposed that an algorithm to identify whether a picture contains a bird would take a team of researchers five years. YOLO made that available on a single desktop by 2018, in the sense that I could and did implement training from scratch, personally. A decade after XKCD 1425, you can buy equipment running (heavily stripped-down) equivalents or alternative approaches off the shelf default-on; your cell phone probably does it on someone's server unless you turn cloud functionality it off, and might even then. People who loathe image diffusers love auto-caption assistance that's based around CLIP. Google's default search tool puts an LLM output at the top, and while it was rightfully derided for nearly as year as terrible llama-level output, it's actually gotten good enough in recent months I've started to see anti-AI people use it.
This post used AI translation, because that's default-on for Twitter. I haven't thrown it to ChatGPT or Grok to check whether it's readable or has a coherent theme. Dunno whether it would match my intended them better, or worse, to do so.
Don't read too much into Torvalds' endorsement. The vibe coded python visualizer he talks about is a small helper tool, not the actual project. It's pretty much equivalent to using vibe coding to write scripts and such (where I don't think anyone has disputed that LLMs can be useful).
As a domain expert in that subfield I find it amusing that for all his "programming guruness", the actual meat and potatoes of that project is a combination of what you'd have learnt on a masters level DSP course in the late 70s / early 80s (including the same mistakes students typically do) combined with imitating the state of the art 50 years ago (https://en.wikipedia.org/wiki/Eventide,_Inc#H910_Harmonizer) on much better hardware. Or to put it another way, when Torvalds is taken out of his comfort zone and field of expertise, he's roughly the equivalent of a third year university student. This is why I think anyone claiming universities are useless is full of shit. "Thinking really hard" (to paraphrase Yud) won't give you the necessary theoretical underpinnings that are required to even realize what you're missing, nevermind do something useful in a whole lot of fields.
Incidentally if someone wants to make an actually decently performing pitch shifter, you could do worse than start with this paper which has a rather good explanation of the basic method as well as some significant improvements to quality compared to most earlier publications that are easy to find and read (it'll still sound warbly on polyphonic input but good sounding realtime polyphonic pitch shifting has to this date been accomplished by only three manufacturers that I know of).
To be fair, my feeling is that even Big Yud does not think he can solve the big problems of mankind by just thinking really hard about it. Likely he knows that some questions like P==NP or finding a theory of everything are basically graveyards of geniuses who had the maximum amount of domain expertise, and did not spend a month trying to solve either without domain expertise.
His arguable successes writing fan-fiction and raising the sanity waterline were in fields where there were still relatively low-hanging fruits to be picked by a very smart autodidact willing to invest a lot of time. The field he finally went into was one which was just not a thing before he popularized it, and my understanding is that he picked it for practical relevance.
Or are you arguing referring to him claiming that AIs would solve problems by "thinking really hard"?
I would argue that measured as performance divided by training, LLMs today are stupid to a level which would let a human claim disability benefits. If you expose an LLM to as much English as a 10yo got in his life so far, you will not end up with a LLM which will be as fluent as a 10yo, but likely less fluent than a 1yo. Similarly, sending an IQ 110 individual to university for a decade to study DSP for eight hours a day would still expose them only to the tiniest fraction of textbooks and papers which an LLM had in its training data. Yet they would handily outperform the LLM.
The advantage current LLMs have is not that they are intrinsically smarter than humans. Instead, it is that while they need much more training material than humans, it can be scaled up just by burning ever-increasing amounts of money. Think 'an immortal IQ 90 guy with an unlimited time turner and no memory loss'. The LLMs performance for DSP may be subpar, but it will offer a similar performance over virtually every domain from OS/2 security to bird symbolism in medieval Polish poetry.
If AIs will ever have an intrinsic intelligence advantage over humans, that would be much closer to what Yudkowsky expects. I think that you require a huge intelligence advantage to outperform domain knowledge. The IQ 140 mathematician whom nobody teaches to count will probably not discover any mathematics not known to a society who has spent a few 10k person-years of IQ 100 mathematicians on the subject. One would need to make the gap more extreme, perhaps she would discover more than any IQ 80 mathematicians would ever discover.
Of course, there is also no reason to suppose that an AI system will not have the better part of the collective knowledge of humanity at the tips of its tentacles. We are blasting the damn things with it during training so that the tiniest amount sticks, after all. Even if we did not, once it has access to the internet, it can basically become a domain expert on any subject it cares about. Even if it avoids libgen and the like, just reading what is available on the non-copyright-infringing open web is enough to get anyone with enough time and brains pretty close to the cutting edge, certainly close enough that it could use its intelligence advantage to make original discoveries.
Ramanujan did.
From what I recall reading about him, he most certainly did not independently discover integers. I think that he had access to some math textbooks fairly early. Now, I will grant you that he got much further with the sparse resources he had than a billion other people would have in his place, but in a world where nobody taught him to count, he would be busy inventing a sufficient notation for numbers and the like.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
That's fair, although I'll caveat that assuming university degrees are good designation of skill and that a skilled user with AI ends up "roughly the equivalent of a third year university student", that's still praising with faint damns.
That's a very interesting paper to read, I'll admit, though.
The joke is of course that DSP is typically introduced in the third year (or at least was around here). Ie. Linus (widely regarded a brilliant programmer and with a masters degree in CS) managed roughly as well as a student after an introduction to DSP course. I don't think he actually used AI for the C code, or at least it wouldn't make any sense given how there's a dearth of good training material (*) and the programming part itself is incredibly easy for any competent C or C++ programmer who doesn't need particularly optimal code (as Linus outright mentions in the repo). The trick is knowing what DSP algorithms to use and how and why the textbook ones are flawed (or outright bad in many cases).
*: There's a site called musicdsp.org which is a somewhat prominent site with pseudo- and sourcecode snippets of all sorts of audio DSP algorithms. They are also almost all from subtly flawed to horribly bad and and a layman (ie. someone who hasn't studied DSP theory) will have a hard time understanding how and where. Thus it's quite ironically almost exactly what you'd get if someone time traveled into the early 2000s and established a site specifically dedicated to poisoning future AI training on that subtopic (with a fair bit of success, I'd say given how the alternatives are either bits here and there or actual books / papers with math instead of code that can be copy pasted). Way back in the day I went from "Hmm, this looks kinda nifty" to "OMG, everyone there is a goddamn moron" in the course of just one year when I started DSP studies.
University degree by itself is no guarantee (as I found out to my pain during a couple of group work sessions when I had to do everything myself when everybody else was so incompetent) but there are many hard science / engineering fields where more or less everyone competent has a degree in a closely related subject, a maths / physics degree with significant self study or are polymaths with near genius level intellect. So in practise a university degree is a requirement to be any good at them. Programming just happens to be a very notorious exception to that (case in point, I only took a handful of programming classes in university and have been making my living for the last 25 years mostly as a C++ programmer in either DSP or embedded systems).
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
I wanted to write a post about some of these events, specifically the change in attitude for the titans of industry like Linus Torvalds and Terence Tao. I'm no programmer, but I like to peer over their shoulders, I know enough to find profoundly disorienting, seeing the creator of Linux, a man whose reputation for code quality involves tearing strips off people for minor whitespace violations, admit to vibe-coding with an LLM.
Torvalds and Tao are as close to gods as you can get in their respective fields. If they're deriving clear utility from using AI in their spheres, then anyone who claims that the tools are useless really ought to acknowledge the severe Skill Issue on display. It's one thing for a concept artist on Twitter to complain about the soul of art. It is quite another for a Fields Medalist to shrug and say, "Actually, this machine is helpful."
Fortunately, people who actually claim that LLMs are entirely useless are becoming rare these days. The goalposts have shifted with such velocity that they've undergone a redshift. We've moved rapidly from "it can't do the thing" to "it does the thing, but it's derivative slop" to "it does the thing expertly, but it uses too much water." The detractors have been more than replaced by those who latch onto both actual issues (electricity use, at least until the grid expands) and utter non-issues to justify their aesthetic distaste.
But I'm tired, boss.
I'm sick of winning, or at least of being right. There's little satisfaction to be had about predicting the sharks in the water when I'm treading that same water with the rest of you. I look at the examples in the OP, like the cancelled light novel or the fake pop star, and I don't see a resistance holding the line. I see a series of retreating actions. Not even particularly dignified ones.
Ah, the irony of me being about to misattribute this quote to Gandhi, only to be corrected by the dumb bot Google uses for search results. And AI supposedly spreads misinformation. It turns out that the "stochastic parrot" is sometimes better at fact-checking than the human memory.
Unfortunately, having a lower Brier score, while good for the ego, doesn't significantly ameliorate my anxiety regarding my own job, career, and general future. Predicting the avalanche doesn't stop the snow. And who knows, maybe things will plateau at a level that is somehow not catastrophic for human employability or control over the future. We might well be approaching the former today, and certain fields are fucked already. Just ask the translators, or the concept artists at Larian who are now "polishing" placeholder assets that never quite get replaced (and some of the bigger companies, like Activision, use AI wherever they can get away, and don't seem to particularly give a fuck when caught out). Unfortunately, wishing my detractors were correct isn't the same as making them correct. Their track record is worse than mine.
The TEGAKI example is... chef's kiss. Behold! I present a site dedicated to "Hand-drawn only," a digital fortress for the human spirit, explicitly banning generative AI. And how is this fortress built? With Cursor, Claude, and CodeRabbit.
(Everyone wants to automate every job that's not their own, and perhaps even that if nobody else notices. Guess what, chucklefuck? Everyone else feels the same, and that includes your boss.)
To the question "To which tribe shall the gift of AI fall?", the answer is "Mu." The tribes may rally around flags of "AI" and "Anti-AI," but that doesn't actually tell you whether they're using it. It only tells you whether they admit it. We're in a situation where the anti-AI platform is built by AI, presumably because the human developers wanted to save time so they could build their anti-AI platform faster. This is the Moloch trap in a nutshell, clamped around your nuts. You can hate the tool, but if the tool lets your competitor (or your own development team) move twice as fast, you will use the tool.
We are currently in the frog-boiling phase of AI adoption. Even normies get use out of the tools, and if they happen to live under a rock, they have it shoved down their throats. It's on YouTube, it's consuming TikTok and Instagram, it's on the damn news every other day. It's in your homework, it's in the emails you receive, it's you double checking your prescription and asking ChatGPT to explain the funny magic words because your doctor (me, hypothetically) was too busy typing notes into an Epic system designed by sadists to explain the side effects of Sertraline in detail.
To the extent that it is helpful, and not misleading, to imagine the story of the world to have a genre: science fiction won. We spent decades arguing about whether strong AI was possible, whether computers could be creative, whether the Chinese Room argument held water. The universe looked at our philosophical debates and dropped a several trillion parameter model on our heads.
The only question left is the sub-genre.
Are we heading for the outcome where we become solar-punks with a Dyson swarm, leveraging our new alien intelligences to fix the climate and solve the Riemann Hypothesis? Or are we barrelling toward a cyberpunk dystopia with a Dyson swarm, where the rich have Omni-sapients in their pockets while the rest of us scrape by in the ruins of the creative economy, generating training data for a credit? Or perhaps we are the lucky denizens of a Fully Automated Luxury Space Commune with optional homosexuality (but mandatory Dyson swarms)?
(I've left out the very real possibility of human extinction. Don't worry, the swarm didn't go anywhere.)
The TEGAKI example suggests the middle path is most likely, at least for a few years (and the "middle" would have been ridiculous scifi a decade back). A world where we loudly proclaim our purity while quietly outsourcing the heavy lifting to the machine. We'll ban AI art while using AI to build the ban-hammer. We'll mock the "slop" while reading AI summaries of the news. We'll claim superiority over the machine right up until the moment it politely corrects our Gandhi quotes and writes the Linux kernel better than we can.
I used to think my willingness to embrace these tools gave me an edge, a way to stay ahead of the curve. Now I suspect it just means I'll be the first one to realize when the curve has become a vertical wall.
Yeah. There's a mirror to the post here I wrote on tumblr, steelmanning the concerns with AI use for that audience, and it's... concerning how many are already applicable even if zero further progress with AI features and capabilities happen.
(And, uh, that I used AI-assisted doctors as an example, albeit of the form "Amazon-East goes down and suddenly your doctor doesn’t know how to check the side effects for a specific medicine.")
Very much agreed. It's probably useful to notice where the criticisms of AI output have moved from errors in basic functionality to lack of Vision or discernment of domain expertise, but it's not very comforting for someone without that vision themselves.
And then it becomes near-certain that progress isn't going to stop here.
More options
Context Copy link
I wouldn't go that far, but in general I agree. (I am thinking of the AI systems that currently stand in the way between me and a live human customer service representative who might actually be able to help me. So annoying.)
As you say, it's kind of a Moloch situation now in which AI offers enough benefits that there is basically a race to use it maximally.
More options
Context Copy link
More options
Context Copy link
Like the Maori and the Moriori debating whether to be "pro-gun" or "anti-gun".
More options
Context Copy link
AIs can add fairly complex features in fairly large existing codebases. I've tried my hand at 'gamedev but I can't code' and it does work, even for more complex things like wargame AI where it needs to move a bunch of units on a 2D map, consider relative strength and threat, manage pathfinding.
The real issue is making stuff that's actually good. If you left AI to its own devices, it'll produce extremely generic concepts, functional but uninspired. Just try it, tell the AI to one-shot a game of some kind. You'll get the exact median in all areas, boring mechanics, boring opponents, boring kinds of variety.
Human level intelligence has been achieved and achieved some time ago IMO. What's needed is wisdom, extremely long time-horizons and better vision/spatial ability. All of that probably comes with more intelligence but it's subtly different from 'can solve even more bafflingly difficult mathematical tasks' intelligence. In the real world, the mathematics is not the hard part. It's using the mathematics, the physics, the chemistry, as just one part of a long process to create a useful product or technology.
I was speaking with a senior dev at AWS earlier today about AI and coding. Amazon has its own tools for both devs and corp staff and doesn't allow outside tools for anything but simple questions who's output will never be in a work product. He said that one of the bigger surprises for him was how many devs, especially the more junior ones, have adopted using AI for a lot of their coding questions and drastically reduced or stopped using the popular developer forums like StackOverflow for research. He asked a few of them what the main appeal of asking the AI was and they said, even though the AI is often wrong and you have to double check all its answers, it never insults you for knowing less than it does or mocks you for trying to learn something, which was an extrememly common experience on the forums. My own work involves some AI interaction intent-verification where my teams review AI-human interactions for new products to judge how well it captures the person's intentions and the quality of responses. From my POV one of the best things about these AI bot tools it they can't (deliberately) lie, automatically making them better than I'd estimate 75% of the people they will likely replace, who lie constantly.
Yeah it's great for that. You can ask really detailed specific questions, supply lots of context and then ask follow up questions, in real time. Whereas on a forum you're waiting and waiting...
https://stackoverflow.blog/2018/07/10/welcome-wagon-classifying-comments-on-stack-overflow/
To be fair, most humans answering questions on stack overflow are unpaid volunteers. (Sure, a few are Godharding their own score for professional reasons, but the most efficient way of doing this is to plagiarize answers from other people to similar questions.)
Giving advice to people who are advise-resistant can be very frustrating even when you are getting paid for it. Sometimes it emerges that the advice-seeker is planning to do the equivalent of adding helium balloons to a tank to build an air superiority platform, and after a lengthy discussion your only impact is that he is now planning to use hydrogen instead to counter your argument about helium prices.
For me, 99% of the value of stack overflow is that a google search redirects me to a question someone has asked years ago. While it sometimes happen that I try to communicate on IRC, github, SO or the like with random strangers to solve a problem I am having, that is basically an admission of defeat, and I tend not to surrender easily.
The tone on SO should reflect that >90% of the value provided by the site is to passive readers. If you ask a new, interesting, relevant question that is providing a tremendous service to the readers, same as answering such a question. If you are asking a question which was already answered five times because you did not bother to google first, you are wasting everyone's time. If your code or design is shit, then anyone pointing that out is providing a valuable service to the community, given that the question will mainly be read by people with a similar knowledge level.
With IRC (or discord?), this calculation is different, because IRC logs are typically not google-indexed. (Or at least I can not remember finding the answer to a technical question in an IRC log not explicitly generated by a human, ever.) There is not really an audience who is the main beneficiary of the interaction.
More options
Context Copy link
More options
Context Copy link
The thing about stackoverflow is most of the time I'm doing a web search and that brings me to a Stackoverflow link. I check the link and it's not the same question I have, it's something superficially similar. That's the same thing AI tends to do, so it's a wash anyway.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
Reading over your prediction, are you certain that it was wrong, or that you wildly underestimated how much money people were willing to throw at the problem despite the quadratic curve? My personal prediction was that we'd see moderate efficiency gains that would help, but I never thought we'd wrap up over a full percent of the US GDP in it.
It'd be a convenient dodge, but even if I were willing to take it, it'd probably would still be wrong. In particular, I was under the impression that no amount of money could flatten the quadrilateral explosion in regards to Attention and Context, and it turned out to not only be doable, but doable in forms that could run on a single computer in my house. (Indeed, at 30B, a single computer in my house can go up to 1m tokens), aka off by 20x my 50k token estimate. It's not free and the software development came from money as much as it did obsessives poking at theory, but it's not like ChatGPT solved it just primarily or even predominately by throwing GPU cycles at the thing.
More options
Context Copy link
More options
Context Copy link
AI artistic successes are indicative of survivorship bias. The way their creators operate is by spamming vast amounts of works and seeing what sticks. Through quirks of fate, a few of them end up successful. This business model is probably short lived, though, as the very spam it relies on degenerates the platforms necessary for their proliferation, so that user interest will eventually decline. Already we’re seeing sites like Deviant Art and Literotica killed off by AI spam. AI will kill off markets rather than improve them.
I agree that's kinda the track we are on at the moment. I really enjoy listening to video essays on YouTube and there has been a noticeable decline in quality in just the last 6 months.
That being said, it's difficult to say what will happen. For example, I tend to rely on the YouTube algorithm to suggest videos for me. But there are surely websites out there which recommend quality YouTube channels and videos.
Of course, it's possible that before long, AI-generated content will be more interesting and engaging than human-created content.
The Youtube algorithm has gotten much worse for me lately, and the thumbnails, specifically, have become terrible even on good channels, I assume for SEO reasons. I might actually start asking Gemini for recommendations more instead.
There's a 'De-arrow' extension that tries to de-clickbait popular videos' thumbnail and title.
More options
Context Copy link
This is because such thumbnails generate significantly more clicks even on quality channels. I blame mobile users who can only see a tiny thumbnail so anything "surprising" sticks out.
More options
Context Copy link
More options
Context Copy link
I don' t think the guy has any clue whatsoever about how uninspired most of the stuff that's 'killed off' is tbh.
More options
Context Copy link
More options
Context Copy link
What's happening? Or I guess as that's pretty obvious, what is this doing to the user base and how are moderators/the sites dealing with it?
For DeviantArt specifically, you had an environment where a sizable portion of site monetization came through sales of merch derived from uploaded content or subscriptions to access gated content, along with a discovery algorithm that largely favored the firehose view as much or more than word of mouth. This had its problems even before AI hit, both the obvious (commissions are much more favorable for newer artists and were badly supported), and the more subtle (minimum and maximum prices made more compliance sense than business sense), but AI gen drew the more serious contradictions to the fore.
With Stable Diffusion, it was possible to just flood thousands of images a week, covering a variety of different common tags, per account, and spammers could make a lot of accounts. Only a couple results would ever break out, and maybe a handful would make any sales period, but that's all it really took for some scammers to find it worthwhile. Contrary to popular belief, I don't think that a majority of submissions were AIgen, even at the height of the initial rush, but a large enough portion were that the firehose view was pretty regularly blasted with repeated pages of AIgen. Worse, there was a lot of suspicion that works not tagged as AI were AI, and the more paranoid were sure >25% were AIgen.
DeviantArt had anti-flooding rules, but were slow to bring enforcement against anything but the clearest spammer. Presumably this had some impact on how much payout the scammers could get, and maybe discouraged some of them, but it left the enforcement invisible and didn't solve the firehose interference problem.
I think the death of the platform is overstated, but it's definitely not favored in the way it was at the start of the COVID era. Some of that's downstream of other related issues: DeviantArt integrated an AIgen capability called DreamUp, and while there's probably some way they could have sold it without optimizing for pissing off anti-AI artists (such as helping fit creator artwork to specific merch categories, having automatic eligibility disabled). They definitely didn't, so now it's mostly known as a way that DeviantArt has promoted some really slop-focused artists like Isaris-AI (not linking because it stretches the definition of 'bikini', >15 submissions/day) or mikonotai (cw: tits in not very convincing armor, >7 submissions/day).
I don't have much idea of what happened at Literotica.
More options
Context Copy link
Meanwhile Pixiv embraced AI and is basically a unifans export vehicle. The nobility of the author is secondary to being able to spam endless niche content users actually care for. Unless you're the top tier of artists whose work itself is stylistically unique enough to be differentiatable but not legacy enough to be prompted in, AI will generate whatever the fuck you want. Deviantart Literotica AO3 Wattpad etc survive on fulfilling authorial actualization and paid commissions, both of which were focused on restricted fetishistic content and not creator skill. AI is killing off the lower tier skill tree of this band of creators, and I'm not sure its a bad thing.
It's going to kill off the livelihoods and acclaim for any artist who operates on any site susceptible to AI spam (which includes mainstream ones like Amazon), as the spam will make it impossible for new artists to attract notice and will even make it difficult for established artists to attract new fans. The AI spam doesn't work by creating equal or superior products, it works by simply existing in vast quantities and being hard to distinguish at a casual glance from legitimate products. It's a form of counterfeit goods when used this way.
Furthermore, it kills off a sort of broader cultural enthusiasm for art which exists and accounts for much of our society's interest in it. Fed by AI content mills and their inferior simplistic content, you might be left with satisfied degenerates who don't care about complexity or meaning and are wholly content with endless repetitive images of their favorite anime characters or whatever, but you won't have the kind of cultural underpinnings that sustains either fandoms, forum media discussions, critical appreciations, or anything else that makes art socially engaging.
This will in turn kill off the production of any sort of non-hyper-commodified art. Who wants to put effort into things if no one's A) going to notice or buy it, or B) even possess the cultural capability of caring?
It is a form of inferior goods, but not counterfeit goods unless it claims to be that which it is displacing.
In addition to many spammers putting AI-gen up on websites that specifically prohibit AI-gen, or putting AI-gen up on websites that require disclosing AI-gen without doing so, there's a very annoying gimmick where people will claim to either have produced something themselves or (more rarely) by AIgen, and then running scraped works through just enough of an img2img process to beat anti-duplication algorithms.
This isn't a counterfeit of this in the sense that a fake dollar is a counterfeit of a real one, but it's closer to counterfeit in the sense that a book getting cloned by a scummy print-on-demand shop is counterfeit, and even closer to a book on bookbinding from a scummy print-on-demand shop. Even though it's advertised as AIgen you could produce with the tools on that site, it's still making promises it couldn't cash out: you couldn't make this sort of image using the tool they presented as part of their site capabilities, because it didn't support img2img.
(It's also substandard, but if you could imagine a non-crap output...)
More options
Context Copy link
It does claim to be that which it is displacing, though. There generally isn't open acknowledgement that AI goods are made by AI, and many sellers attempt to actively claim otherwise.
In either case, I'd say there's currently an implicit assumption by many buyers that, when they're purchasing a book, say, they're buying something that an intelligent mind constructed using skill and artifice (with plot twists, character arcs, and so forth), and not something that reads beautifully on the first page but never builds up to anything or has anything to say. AI's utility in this regard is its ability to both impersonate more meaningfully crafted human products and to exploit the sort of assumptions that book customers have built up through former habits.
The result will be the death of those 'former habits', as book customers do not gain the same pleasant experiences from their current purchasing habits, insofar as they inadvertently purchase AI products, and so the market will shrink and utility will be destroyed. If AI products were merely inferior, they could simply be ignored and filtered by such customers. It is their ability to mimic which makes them destructive. They can inhabit certain aspects of outer forms but not provide the same deeper experiences.
Maybe the case is less true for AI drawings, which are more of a what-you-see-is-what-you-get affair, in which case, no, it wouldn't count as a counterfeit to my mind. Unless the buyer was hopeful for some sort of engagement with an actual human that they weren't actually getting, or if they thought they were buying a more complex work that could be intensively studied to extract meanings which weren't immediately obvious, only to eventually realize it's an AI gestalt of several other works which only mimics their qualities superficially, say.
Which seems like a probable enough outcome.
That assumption (which was never true before AI) does not make the goods counterfeit.
More options
Context Copy link
I agree that in a lot of cases, there is AI-generated content where the authorship is concealed.
Isnt the whole genesis of this thread the converse whereby authors conceal their AI generated content?
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
Likewise many stock photo sites are now so full of AI slop that they're becoming useless unless they have a search feature to only search pre-2022 era photos. Otherwise the results are full of the same generic third rate obviously AI photos.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
This... varies pretty heavily by area and focus. The Furry Diffusion discord has some anti-spamming measures and a general ethos focused toward quality, and as a result it's able to keep the 'floor' pretty high and higher-upvoted images are generally pretty high-quality too. They're not all good, and even the greats aren't perfect, but the degree of intentionality that can be brought forward is far greater than most people expect.
That depends on both moderation that may scale in the face of a genuinely infinite slop machine and relatively low stakes (and, frankly, monomania), but it's at least pointing to ways AI creators can operate outside of full spam mode.
Not being an art expert, I can’t judge those images too deeply. One thing that stands out to me though is how compositionally simple those examples are. They seem to all consist of one character in the foreground and then some kind of dramatic stylistic background. My own experiences with AI image generation is that it’s very difficult to get the prompt engine to orchestrate more than just one or two characters, so that this sort of simple approach seems like it is probably the best that current AI is capable of. To me, it doesn’t seem like a rich tool for self expression.
This is why I consider 99% of examples of any new AI image generator basically useless since they're only exhibiting the ultra-easy mode of one or two clear concepts where the layout, position, exact pose and such are irrelevant.
More options
Context Copy link
That's fair. There are some models that allow more specific control prompt-only of multicharacter composition, like Whisk, Nano Banana, and Qwen, but they have tradeoffs and tend to give 'worse' output quality if used as the only or final part of a workflow. In-painting can give phenomenal amounts of control for very complex character layouts (or background layouts), but at the cost of a lot of tedious work (cw: 9mb video file). There's been similar efforts using related technologies for comics, loresheets, game environments, and ultra-complex characters (in the furry fandom, usually things like cyborgs and complex hybrids).
Which does give more space for self-expression, but it's not going to have the volume to be visible in a DeviantArt firehose view.
I've done my time with Stable Diffusion, from the closed alpha to a local instance running on my pc.
Dedicated image models, or at least pure diffusion ones, are dead. Nano Banana does just about everything I need. If I was anal about the drop in resolution, I'd find a pirate copy of Photoshop and stitch it together myself, I'm sure you can work around it by feeding crops into NB and trusting they'll align.
All of the fancy pose tools like ControlNet are obsolete. You can just throw style and pose references at the LLM and it'll figure it out.
I suppose they might have niche utility when creating a large, highly detailed composition, but the pain is genuinely not worth it unless you absolutely must have that.
Yeah, Nano Banana (and Whisk) are stupidly powerful, and don't really seem to have a local or open-source competitor yet. Qwen Image /Image Edit can kinda work on similar principles, and can do some level of scene composition or pose transfer, but it's limited and gets pretty ugly. A number of furry diffusion users start from Nano Banana prompting, then do the final work with a local image model (whether for upscaling, changes in content, or NSFW).
I dunno that I'd call ControlNet obsolete, but that may reflect my own unfamiliarity with Nano Banana (and not using the paid version) as much as anything deeper.
More options
Context Copy link
More options
Context Copy link
This is actually a very heartening video! It shows that you can make a complex scene that doesn't have this PonyXL house style. How do AI artists deal with preserving character details from image to image? It seems to me this is even more important for furry art (various fur patterns must be harder to reproduce correctly than "black hair, pixie cut").
[cw: lots of furry images. nothing involving nudity in any sense but the Donald Duck or swimsuit sense, but probably not something you'd want to explain to your boss]
It depends a lot on what you're aiming for. It's possible to get text-only prompts that retain fairly good consistency of a character. Some of that's because the character itself is pretty 'standard', although they also have a number of potential faults (eg, border collie with a floppy ear and a spot around one eye seems easy, but a lot of models struggle with the "my left or your left" problem). And these can require pretty serious levels of detail and description, much of which wouldn't be obvious to non-artists.
If you've already got a single piece with the character and want a second one in an entirely different context, tools with more semantic understanding focused around transfer like Qwen Image Edit, Nano Banana, and Whisk can do that surprisingly well (albeit generally on the cloud and censored: afaik, only Qwen Image Edit has a local mode). I'd expect some multimodal LLMs could do something similar, but I've only really tried GLM-4.6V for local multimode and never got anything particularly exciting from it.
For one-offs with more specific or complex markings or fur patterns, especially around the face or hands/paws, you're usually going to see a lot of inpainting. The threshold where that becomes necessary can be surprisingly low: this guy seems trivial at first glance, but since it's not supposed to have a few tells from real maned wolves that's often something he had to tweak aggressively, and the four markings on the forehead are really not something most AIgen wants to do as part of a facial structure, so he'd often be loading up krita to help do inpainting. It's still not 'real' artwork, but it can get fuzzier on the edges.
If you plan to reuse the character, doing a few works with inpainting, traditional media, 3d modeling software, or some combination of the above, then building a LoRA tends to be the most effective. A good LoRA takes a lot of effort, but it can be done with a surprisingly small number of reference images and maintain a lot of detail or handle very strange layouts.
For an example, I'll use uverage. He's an avali-wolf hybrid, so he's got a lot of unusual features (the four ears are intentional, the ring marks around the ears and thighs are not standard, and his tail is probably derived from another VRchat species) and while avali are popular enough (6k e621 images) as fictional species go enough he's probably not the first avali-wolf, there's not exactly a surfeit of non-AI training data that matches what conclusion this particular aiGenner came up with. Yet the LoRA can carry markings and physical characteristics across styles, perceived 'medium', or even transfer markings to gender or to other species.
It's far from perfect. Notably, the arm feathers and crest tend to come and go randomly, and the LoRA seems to be messing with the finger-and-toes count. That might be an intentional stylistic decision, but probably not. And LoRAs do have costs: poorly trained LoRA can degrade image quality, and they seldom scale above three or four LoRA in one generation (either text2img or inpainting) before the models tend to just go nuts. But it's the sort of thing that's practically doable at small scales by individuals without too autistic a level of focus.
That said, I will caveat that enough furries are faceblind enough, or otherwise tend to identify characters more by mood, dress (as little as that might be), and large high-contrast markings. I don't know how well the same approach would transfer to realistic or even anime-like humans, especially for an audience with better perception about microexpression or sensitivity to smaller errors; the few examples I'm familiar with tend to be side characters in content I'm not gonna link here.
More options
Context Copy link
Nano Banana or GPT Image are perfectly capable of ingesting reference images of entirely novel characters, and then just placing them idomatically in an entirely new context. It's as simple as uploading the image(s) and asking it to transfer the character over. In the old days of 2023, you'd have to futz around fine-tuning Stable Diffusion to get far worse results.
More options
Context Copy link
I think by and large they are terrible at it and don't. There are a few different techniques that claim to achieve this, but as someone who follows this closely it's all still fairly bad. By far one of the biggest remaining hurdles of mass commercial use.
Matching Eye colour hair colour, clothes etc are doable with stuff like retraining the model, a reference or prompting with a well known actor/figure
God forbid you try to recreate a character that passes the filter of someone who's not faceblind
Want to bet? I’ll wager up to US $500 that I can produce a 30 second video with a consistent, recognizable character using Veo (either Flow interface or API, your choice). Max Veo length is 8 seconds so that’s keeping consistency across 4 generations. We can do cuts to scenes within one gen if you want.
Want to agree on details? This offer is open to anyone.
The bar for me is not that it's recognizably consistent. It's actual consistence. For something like this to cross the commercial viability threshold stuff needs to stay on model.
The character needs to stay consistent in different lighting conditions, angles and FOVs.
Finally it needs to be able to handle unique appearences, not average pretty faces and clothes.
The issue isn't that it's impossible to make a video of a character from an image be consistent with that image. Although in my opinion we're still not there. The difficulty arrises from the fact that such a video will inevitably have to conjure up new details in the process. Keeping the newly created information consistent with the next generated clip gets exponentially harder with each new clip and required context. Similar to how LLMs fail if the context is long enough.
I doubt you can make something like bill gates wearing tiger face paint and a floppy sleeping cap from a flat front shot, to an over the shoulder partial view, to a side view without messing up the direction of the flop of the cap or the position/amount of tiger stripes in the make-up.
Not going to bet money on it because I'm sure with enough tries it's doable, I'm just illustrating a point that the amount of stripes and flops or whatever is essentially the same as subtle facial features like the angle of the jaw or the tilt of the eyes.
The technology is fundamentally just not designed for this sort of thing. There's tons of workarounds and it will still be very impactful, you can work within the constraint to achieve amazing stuff, but the constraints are still there.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
Human artistic successes are indicative of survivorship bias. AI just makes this more visible because the productivity is so much higher.
It also amplifies the effect through the amplified productivity. That is, you can achieve greater success with a lower mean quality, because instead of having a thousand humans write a thousand works and then pick the best one, you can write ten million AI works and then pick the best one, allowing you to select more standard deviations up. Which means that there will be literal millions of AI slop work of very low average quality just in the hope that one will rise to the top.
This makes discovery a lot harder and waste more time from pioneers reading slop in order to find the good stuff.
I’m not a ‘math wizard’, but something about this seems off. Shakespeare didn’t write one hundred plays and then choose the best few dozen to publish. He developed his playwriting ability until he was at a skill level to consistently put out good content. If AI lacks the underlying skills to create good works, then should we expect even one in a trillion of its works to be as good as Macbeth, or should we regard the creation of such a thing as physically impossible unless underlying conditions are met? It seems like it’s less a matter of statistical probability than physical possibility.
Survivorship and selection bias works on the population level as well as the individual work level. How many hundreds or thousands of playwrights existed in Shakespeare's time? And yet most are forgotten, while the best of the best (Shakespeare's works) are what are remembered and enjoyed and studied.
Also, there definitely is variation within an individual author's works. How much time and effort do people spend studying "Two Gentlemen of Verona"? Is it actually a good work? Personally I haven't read it, but given how little it's talked about or ranked on people's lists, my guess is that it's mid and the only reason anyone ever talks about it at all is because Shakespeare is famous for his other plays. That is, Shakespeare wrote 38 plays and, while his skill was well above average, and therefore his average work is higher than the average play, they're not all Hamlet. But one of them was. He didn't write a hundred plays and then only publish the best, he wrote 38 and then published them all and then got famous for the best few (which in turn drove interest in the rest above what they actually deserve on their own merits).
In-so-far as AI is likely to vary less in "author" talent since whatever the most cutting edge models are will be widely copied, we should expect less variance in the quality of individual works. But there will still be plenty of variation, especially as people get better at finding the right prompts and fine-tuning to create different deliberate artistic styles (and drop that stupid em-dash reliance).
I tentatively agree that there are limits to this. If you took AI from 5 years ago there is no way it would ever produce anything publishably good. If you take AI from today I don't think it could ever reach the upper tier of literature like Shakespeare or Terry Pratchett. However this statistical shotgun approach still allows one to reach above their station. But the top 1% of AI work today might be able reach Twilight levels, and if each of those has a 1 in million chance of going viral and being the next Twilight, then you only need to publish a hundred million of them and hope you get lucky. Clearly we've observed that you don't need to be Shakespeare in order to get rich, its as much about catching the public interest and catering to (and being noticed by) the right audience as it is about objective quality, and that's much more a numbers game.
I do think that AI lacks the proper level of coherence and long-term vision to properly appeal to a targeted audience the way something like Twilight or Harry Potter does. But a human curator (or possibly additional specialized AI storyboard support) could probably pick up the slack there (although at that point it's not quite the shotgun approach, more of a compromise between AI slopping and human authorship, and mixes the costs and benefits of both)
Shakespeare and his contemporaries had to pay an upfront cost that was significantly higher. They had to write (already a rare skill) a manuscript and had convince at least one theatre manager to read their work. This means their innate skill had to be high enough that their first (or second, or third if they were persistent) play was of sufficient quality already.
A modern ShAIkespeare can produce and publish a new play every weekend. We need Lord StrAInge's Men, a troupe of AIs that can read, review and dismiss AI slop just as quickly as it's written instead of relying on avid human readers.
An AI that can accurately identify and dismiss slop is 90% of the way towards producing quality content, since you could just build the generative AI with that skill built in (and train them on it).
Which is to say, maybe in 10 years this will be a mostly non-issue. If they reach the point where they can generate thousands of genuinely high quality and entertaining stories, I'll happily consume the content. I think "human authorship" as a background principle is overrated. It has some value, but that value is overrated in comparison to the actual inherent value of the work. The problem with slop is that it's not very good, regardless of whether it's generated by humans or AI. Once it's good then we're good.
Maybe? Or it might be easier to do as a separate model. Hell, it's possible even dumber solutions might be more readily made.
More options
Context Copy link
Not if the process itself is beyond the AI to recreate.
For instance, say that a great movie like A Clockwork Orange was made in part through the theoretical understandings the main actors had developed over their lifetimes for their crafts and used to feed into their decisions of how to act and portray their characters.
Coming up with a similar quality of acting might be impossible through mere observation and mimicry of what works and what doesn't. The AI has an intuition for what sorts of things generally go together, but it doesn't use, among other things, underlying theoretical know-how to construct its outputs.
My current assessment is that there's a low ceiling for how far AI 'thinking' can take the quality of its output, particularly regarding the complexity of what it's attempting to do. Projects that require a layered approach of various theories and techniques seem like they're fundamentally beyond AI. The more systems that need to come together to create a work, the more exponentially difficult it becomes for a pattern-finder to match its quality. The pattern-finder needs to become capable of wielding tools, systems, theories in its thinking in order to up its game past a certain point.
I've heard people say before, in the context of AI art, that humans are essentially just 'pattern finders', too, and so are creatively indistinguishable from AI. But I think this is wrong: it ignores external tools humans use to structure and create their work, such as theories and techniques, which cumulatively take the load off of them having to conceive everything in a fit of genius. I think this is the primary reason AI, despite its 'brilliance' as a search engine or generalist explainer, is so lacking in certain other regards. It's due to the total reliance of its 'cognition' on what, compared to humans, would be more like a single sub-process.
More options
Context Copy link
Yeah, once it's good then it's good. The problem with AI content is that it can be produced by people that have no taste. Kinkade's success was one of a kind, but now anyone can create an equally terrible picture that has all the signs of expert craftsmanship. And since there are lots of people that have no taste (or else Kinkade would've died a much poorer man), they can all make and consume terrible art.
I liked Scott's "AI art vs human art" contest and another iteration in 2026 using the same rules would be even harder, but it was rigged: we didn't get random AI art vs random human art (or even randomly sampled from high-rated examples): we got a set of pictures explicitly designed to resemble human-made art. It's like when people failed the Turing test against a chatbot that pretended it was a non-native teenager (this happened before LLMs): the fact that some bots can fool you in specific circumstances doesn't mean you can't complain about the rest of them being chipper hallucinators.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
Some writers, Stephen King comes to mind first, are famous for writing prolifically and then substantially editing down their products. Although maybe not quite that ratio.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
Not true. Human works that find great success usually do so based on their merits as artistic products. AI works that find success usually do so as flukes. Put out millions of AI created light novels and occasionally one of them will slip through some quality filtering service. Their success is predicated on the inability of these services to filter quality 100%, and they enjoy an advantage over the shittiest of human works in this regard based only on the scale of their output.
I've seen enough experiments showing successful art and music are mostly random to think this is definitely not true.
People walk through an art gallery and are asked to rate their favourite pieces. It's like an even split.
When you mix in "experts" to tell people whether the art is good or bad, the random walk disappears and everybody just agrees with them.
You're describing three hundred years of the publishing industry.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
Until last week, Google's AI pro antigravity plan with the annual deal was one of the cheapest ways to get Claude Opus 4.5. Unclear what the impact of their new weekly rate limit is going to be
More options
Context Copy link
My understanding is that this was not actually a prestigious award and may in fact have been done for publicity.
Huh. Fair if true. I saw (and was familiar with) Six One Indie for a couple previous years of their showcases, but I stand to be corrected if you have more detail.
Not 100% sure on that since the source was a reddit comment but a cursory google makes it seem to be the case.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
While the most famous mainstream family of LLM is the one by OpenAi, it is Anthropic which is doing the most interesting theoretical work, and producing novel ways of utilizing this technology. (Ironic, given Anthropic's unorthodox beliefs regarding AI: that censorship is partially for the benefit of the model.) Anthropic focusing its efforts less on making the best artificial friend and assistant, and more on making the best artificial assistant, has produced results expected of specialization: success. As a recent The Atlantic article shows, even normies can now utilize the computing power humanity has created.
More options
Context Copy link
I bet various AI companies are actively sponsoring TEGAKI because they need a clean source of new training data.
It's plausible, but TEGAKI seems... questionably competent enough that it'd be a weird bank shot. Beyond that, a lot of recent image models (Flux, Qwen Image/Edit, Nano Banana) show increasingly strong evidence of (or outright state they are) being trained heavily on synthetic data.
More options
Context Copy link
More options
Context Copy link
Yeah you bring up the point that people will get up in arms about someone else using Generative AI in ways they disagree with. But everyone will have some use of generative AI that they either partake in themselves or approve of generally.
Hence why Pro-AI forces will win in the end. An artist who is up in arms about being out-competed by Slop will probably use an LLM to complete some task they consider 'beneath' them and/or not worth paying money for someone else to do. Its just too useful across too many different tasks.
I do not think we're at the literal cusp of superintelligence... but I do think we've passed a point where the cutting edge LLMs are now smarter/more capable than the median human, even the median American in purely 'mental' tasks.
If this were the culture war of 4-8 years ago, there'd probably be multiple articles about digital blackface given that the AI product is presented as a black woman's voice and image.
Yeah, it's a hard topic, and a scary one. I was considering linking this post from tumblr:
And in some ways, it's a funny and illustrative story, and if AI freezes at exactly this state, I'd expect that we'll see a bunch of people very proud of their predictive prowess. And then it's also a funny and illustrative story, because 'can compete with you for every skill but your one or two specific areas of focus' describes the entire process of employing skilled labor everywhere.
Thing is I think they are all mostly right (for now).
AI can get you a very good rough draft that reads as subject matter expertise once you guide the AI through some adjustments, but is still pretty hard to work with when it comes to the final refining and polishing, the little details here and there, and particular word choice, design, choice, and formatting choices, as well as something like the 'taste' of a true expert judging the finished product.
I could have it draft legal documents that would absolutely pass muster in Court, but don't quite meet my standards for work product I want to sign my name to.
Maybe this say somewhat more about the standards of the practitioners/artists being 'too high' for day to day work, or its the fact that the audiences in question just don't care enough about the quality differential.
For most purposes, if I'm a non-expert relying on the AI to fill in skill gaps, I'll presumably do fine, but still fall short of the output quality of a dedicated team of experts.
More options
Context Copy link
As a teacher, I've given basically the same guidelines to my students: in a writing class, use of AI-generated text is cheating comparable to plagiarism, but students are permitted (as long as they're upfront about it and get my approval first) to use AI for non-text creative projects.
In my case, it's not because I assume that 'jobs I'm unfamiliar with... can be replaced by automation.' It's because the class and the assignment is intended to teach writing skills, not art skills. It's because I am a good writer, and able to teach writing skills; it's because I am an abysmal artist, and not competent to teach art skills. As long as they're dedicating their full effort to the actual class material, I don't mind if they use shortcuts for peripheral tasks. It's not about my estimate of what AI and automation is capable of; it's about my estimation of what I am capable of, and what I expect my students to be capable of.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link