FaceDeer

FaceDeer, 4 months ago

X, apparently.

FaceDeer, 4 months ago

Are you going to somehow reach into my personal computer and remove the software and models from it?

FaceDeer, 4 months ago

Ah, it was the third option, ignorance.

FaceDeer, 4 months ago

If it's horrible and it's also "masquerading" as human art, what does that say about human art?

FaceDeer, 4 months ago

No, I'm just pointing out the common contradiction I see in threads like this, where people argue that AI is both a big threat to "traditional" artists and also that AI is terrible compared to "traditional" artists. It can't really be both.

FaceDeer, 4 months ago

Break it down into chunks and assemble it like Lego.

FaceDeer, 4 months ago

He put quotes around the word "art", which gives me the opposite impression.

FaceDeer, 4 months ago

And yet there's still plenty of traditional restaurants.

Fast food provides a new option. It hasn't destroyed the old. And "terrible" is, once again, in the eye of the beholder - some people like it just fine.

FaceDeer, 5 months ago

No, agnosticism is a whole other issue from any of this. Agnosticism, technically speaking, is the position "it is impossible to know whether a god or gods exist." That's a separate position from "I do not believe that gods exist", "I believe that gods do not exist", and "I believe that gods exist." You can be an agnostic theist or a non-agnostic atheist. They're along two different axes, like the Dungeons and Dragons alignment system with the law/order and good/evil axes.

Unfortunately the term has gained some additional meanings in common parlance, where it can commonly mean "I'm an atheist/theist but I don't want to say that because it gets me in trouble." Or "I'm not sure what I think so I'm going with the option that sounds the most unsure."

It's led to a huge mess when trying to categorize belief systems in polls like this one.

FaceDeer, 5 months ago

It's like the classic "objection!" "On what grounds?" "It's devastating to my case!" Scenario.

Throughout history technology has repeatedly been developed that lets people do things faster than the people currently doing it. That's usually the point of technological progress. Of course the people left behind by that will complain, but that alone is no reason to limit the rest of us who would benefit from the advance.

FaceDeer, 5 months ago

Anyone who enjoys creative things, since they now have access to a lot more of it a lot more easily.

FaceDeer, 5 months ago

Then it's not a threat to professional creatives, is it?

FaceDeer, 5 months ago

How dare people have different priorities than you.

FaceDeer, 5 months ago

You said:

Substantial reproduction of the original work, you can get back substantial portions of the original work from an AI model’s output.

If an AI is trained on a huge number of NYT articles and you're only able to get it to regurgitate one of them, that's not a "substantial portion of the original work." That's a minuscule portion of the original work.

FaceDeer, 5 months ago

It's actually the other way around, Bing does websearches based on what you've asked it and then the answer it generates can incorporate information that was returned by the websearching. This is why you can ask it about current events that weren't in its training data, for example - it looks the information up, puts it into its context, and then generates the response that you see. Sort of like if I asked you to write a paragraph about something that you didn't know about, you'd go look the information up first.

but humans also can differentiate between copyrighted and public works

Not really. Here's a short paragraph about sailboats. Is it copyrighted?

Sailboats, those graceful dancers of the open seas, epitomize the harmonious marriage of nature and human ingenuity. Their billowing sails, like ethereal wings, catch the breath of the wind, propelling them across the endless expanse of the ocean. Each vessel bears the scars of countless journeys, a testament to the resilience of both sailor and ship.

FaceDeer, 5 months ago

I have used it as a collaborator when doing creative work. It's a great brainstorming buddy, and I use it to generate rough drafts of stuff. Usually I use it while developing roleplaying scenarios for TTRPGs I run for my friends. Generative AI is great for illustrating those scenarios, too.

FaceDeer, 5 months ago

As I've been expecting all along, the competition is not going to be between humans and AIs. It'll be between humans that use AIs and those that don't, and the ones that don't will be at a big disadvantage. SAG-AFTRA saw the writing on the wall and wanted to keep its monopoly on what voices are "allowed."

FaceDeer, 5 months ago

It's not as expensive as you might think, there's been a lot of interesting developments lately in making LLMs more resource-efficient.

FaceDeer, 5 months ago

Starship's ITF-2 launch already got well above the Karman line.

FaceDeer, 5 months ago

The booster was never intended to go above the Karman line. Calling that a "failure" is ludicrous.

Also, the orbiter was destroyed by its flight termination package triggering, which is the very definition of an intentional action. The reason it triggered was apparently an oxygen leak that led to the upper stage running out of oxidizer just a few seconds short of achieving orbit, which wasn't according to the flight plan, but this was a test flight so the plan was always "see what happens and fix whatever problems come to light" so that's still not exactly a failure. They got farther than they did on IFT-1.

You are perhaps more used to the NASA way of "testing", which is to exhaustively perfect the rocket before it ever launches and then expect everything to go smoothly during a single shakedown flight before payloads start going up with flight #2. That's not how SpaceX does things.

My prediction for 3 is that again at least part of the craft will blow up below the Karman line.

Given that the booster is never going to cross the Karman line (booster separation happens at 64km), and that the intention is to deliberately ditch the booster in the ocean rather than recover it, you've got quite a conservative prediction there. I honestly can't think of any possible way that this wouldn't happen.

FaceDeer, 6 months ago

Yeah, in discussions like this it's important to put out both the "it's okay if you've got a deep curiousity/desire to be the other gender that you want to explore to see if it leads somewhere more" and the "there's nothing wrong with just having fun exploring other identities or bodies without it being some kind of deep-seated transgender thing." I think the "egg_irl" reaction is sometimes harmful because it ends up pressuring people who really aren't transgender but who would be perfect allies if they weren't ending up feeling annoyed by the whole thing.

FaceDeer, 6 months ago

And it's also perfectly fine to dress however you like in private, too.

FaceDeer, 6 months ago

You may not be aware of just how annoying and downright offensive it is to have eager "helpful" people instantly jumping to "aha, you're trans and I will help you come to terms with that!" When you mention that your roleplaying characters or whatnot play around with various genders. It's probably not quite on the same level as people assuming gay people are just "rebelling" or "going through a phase" or whatever and will just get over it, but I imagine it feels along those same general lines.

There surely are some people who are indeed a metaphorical "egg" just waiting to crack, but everyone should have the right to feel comfortable with themselves regardless. Dismissing those offended feelings as walking on eggshells misses that point.

FaceDeer, 6 months ago

A classic pattern, the "monopoly" provider's service declines steadily over the years because they've got no reason not to be lazy until one day all of a sudden there's a new competitor and they find themselves scrambling to suddenly care again.

Thank you, ChatGPT. :)

FaceDeer, 6 months ago

This was already true years ago after search engines became a thing. The main answers that come to mind for your question are:

• providing novel information that wasn't online before.
• providing information to you that you wouldn't have thought to ask for on your own.

Both of these remain valid and useful reasons for going to a web site even if that web site's content is AI generated.

There's also the matter that "AI generated" is a very broad term. Did someone merely turn an AI loose with a vague instruction to generate some pap to fill a page out with? Or did someone actually provide it with a subject and some information to write about and give the resulting article a read-through to ensure it was good? Did they write a rough draft and just have the AI do the polishing? There's lots of approaches here, some of them much better than others.

FaceDeer, 6 months ago

Or, because humans are biased.

FaceDeer, 6 months ago

I'm reminded of the apocryphal Ghandi quote "first they ignore you, then they laugh at you, then they fight you, then you win." It seems like the general zeitgeist is in between the laugh/fight stages for AI right now.

FaceDeer, 6 months ago

Sounds like basically the same reason I quit editing Wikipedia over a decade ago, so it's not just a generational thing.

FaceDeer, 6 months ago

I know I can edit, I just don't want to any more. The fun got sucked out.

FaceDeer, 6 months ago

And Quebec separatism wasn't its own particular flavour of pre-MAGA MAGA? I think Quebec would have done a fine job of turning it into a shitshow on its own.

FaceDeer, 6 months ago

Last I saw Lightning was pretty much DOA, it's been around for many years and almost nobody's using it. At the time I was checking there was an order of magnitude more activity transferring Bitcoin on Ethereum using WBTC tokens than using Lightning on Bitcoin itself.

FaceDeer, 6 months ago

I recall back when Lightning was first come up with thinking how incredibly hacky a solution it was, full of awkward workarounds for the limitations of Bitcoin's blockchain. A few small changes to the blockchain would have made it so much simpler and more robust, but at that point Bitcoin's immutability had become such a fundamentalist religion that any such changes were absolutely rejected. They wouldn't even fiddle with the block size, let alone consider expanding its scripting capabilities.

Then Ethereum came along with the exact opposite philosophy, it's willing to continue making changes to the foundation layer with the overall goal of making Ethereum more functional for diverse applications. Ever since then it's just been a slow transition of everything useful moving over to Ethereum and Bitcoin becoming ever more insular and obsolete by comparison.

It seems like I haven't thought about Bitcoin in years. When this article came up it took me a moment to shift the mental gears and go "oh yeah, that. I guess it's still around."

FaceDeer, 6 months ago

Yeah. Right now, the cost of a Bitcoin transaction is around $65 US. That price includes all of the expenditures that the miners have made on resources (electricity, water, rental costs for the space they're using, hardware depreciation, etc.), as well as whatever bit of profits it takes to keep miners in business. That puts a cap on whatever environmental impact the transaction is having.

FaceDeer, 6 months ago

Bitcoin is inflationary, it's generating new Bitcoin with every block and issuing that to the miners. That new Bitcoin combines with the transaction fees to pay the miners.

FaceDeer, 6 months ago

Maybe in the long run. However, when you want to actually calculate how much each transaction costs, you need to account for the fact that right now Bitcoin is inflationary. It won't stop issuing new tokens until around 2140 AD, assuming no hard forks happen to modify that issuance strategy in the meantime.

FaceDeer, 6 months ago

It's an overestimate. Right now the total cost per transaction is around 82 US dollars, which at current exchange rates is 75 Euros. That cost covers everything - electricity, rent for the building, salary for staff, taxes, depreciation of mining equipment, and whatever profit is required to keep the miners in business. I don't know what proportion of miner costs actually goes to electricity but I expect it'll likely be much less than 70 Euros.

Perhaps someone got that 700 kWh figure by doing the reverse calculation - looking at how much a transaction cost and then assuming that all goes toward electricity.

FaceDeer, 6 months ago

Yes. If you're wanting to know how many resources mining a transaction takes, that's the value you need to look at. The block reward effectively goes into subsidizing the transaction fees that are being paid.

FaceDeer, 7 months ago

Misleading. Ad blocker installations also rose. This isn't people leaving adblocking, this is people changing to better ones.

FaceDeer, 8 months ago

"By their own admission" doesn't necessarily hold any weight, though. They're not experts, and even if they were experts they can still be wrong. We've got treatments for a variety of psychological disorders these days, with varying efficacy, I don't think it's unreasonable to expect that someday we'll make progress on this sort of thing too.

FaceDeer, 8 months ago

We have the tech right now to shoot people dead. It's just tech, create laws about what's acceptable to do with it and what's not.

FaceDeer, 8 months ago

This thread is about how to regulate AI-generated child porn, that's going to require creating laws too.

What's your proposed alternative to using laws to regulate tech?

FaceDeer, 8 months ago

Well, mainly just one. There are lots of other places in the world that do okay at it.

FaceDeer, 8 months ago

That's a symptom of overfitting, which requires the image to be repeated in the training hundreds or even thousands of times. That generally only happened in earlier image generation models, more "modern" ones ("modern" in this case being measured in months because this is such a fast-developing technology) have much better curation of their training sets to avoid exactly that sort of thing. Nobody wants AI image generators that replicate images from their training sets, what would be the point?

So if you want to find an image model that gives you a close duplicate of an existing image of child abuse, you'll need to find one that was sloppily trained with a training set that included hundreds of duplicates of child abuse imagery. I kind of doubt you'll be able to find one of those.

FaceDeer, 8 months ago

There's a lot of misconceptions about AI image generators in here.

They can indeed generate "truly new images", ask an image generator for an image of something that definitely doesn't exist in its training set and it'll likely be able to come up with something like that for you. Most importantly for purposes of this discussion, you don't actually need to have any images specifically of child abuse in a model's training set in order to train it well enough to produce images of child abuse. Train a model with a bunch of regular porn and a bunch of ordinary images of children and I expect it'll figure out how to make images of children in sexual situations if you ask it to.

This has been known for years. These AIs are capable of "understanding" the things they're trained on and creating novel interpretations of those things.

There was an article recently that showed if you trained many generations of AIs on just the outputs of previous generations you got degraded performance over time, but that's a pretty specific scenario that doesn't match what's being done in real life. In real life synthetic training data (ie, AI-generated training data) can be very useful for expanding the capabilities of AI as long as it's well-curated (humans need to select good outputs and ensure they're described correctly) and ideally has some of the earlier training set's original data included as well.

FaceDeer, 8 months ago

As I said above, though, you don't need to make a model that's specifically "for child porn" in order for it to be able to generate child porn. There are already probably plenty of models that know what children look like and also know what porn looks like, made simply by teaching a model about lots of diverse subjects that happened to include both of those subject areas in them. You can even make new models by merging two existing models together or by adding more training to an existing model, so you wouldn't even need to have those images be part of the same training run.

I obviously haven't ever tried generating child porn, but I fired up my local Stable Diffusion with the Cyberrealistic model and generated a toddler on the moon and a toddler riding a lion. I'm reasonably confident that the model wasn't literally trained with images of toddlers in space suits or toddlers riding large wild predators, it was trained on those concepts separately and was able to figure out for itself how to combine them. Notice how it was able to figure out that a toddler on the moon would be in a space suit and re-proportioned the space suit accordingly, and that a saddle used by a toddler would probably have handlebars (I'm guessing it has a bunch of images of toddlers riding ponies that it got that idea from).

FaceDeer, 8 months ago

Medical textbooks.

FaceDeer, 8 months ago

I think you haven't made much use of image-generating AIs. They're quite capable of reinterpreting images into different styles. A very common use case for me is to draw a sketch of something and then tell an image AI to turn it photorealistic. The "automated collage" approach you describe is simply not how they work, it's a common misconception. Image AIs very much can create imagery of things that weren't explicitly in their training set, they're not just regurgitating pasted-together snippets.

You're also assuming that there are no literal photographs of children's genitals in medical literature. Again, I haven't exactly gone looking, but I'm sure there are some out there. Doctors can't afford to be prudish.

And finally, you can get plenty pornographic without even specifically showing off genitals.

FaceDeer, 8 months ago

Oh, so you mean the photos in the training dataset that violate medical privacy

If they're published in a textbook then they're not private.

will not have given informed consent for those photos to be inserted into an AI model for child porn

Again, an AI model doesn't have to be created specifically for the purpose of child porn in order for it to be able to generate child porn. Most of these AI image models are very general purpose, they can create images of all kinds of things.

We're going in circles here and you're just getting angrier in your responses, I don't think this is headed anywhere useful at this point.