These days, just about anyone with a pulse can fall on a keyboard and make an AI image generator spurt out some kind of vaguely visual content. A lot of it is crap. Some of it’s confusing. But most of all, creators hate it when their hand-crafted works are compared with these digital extrusions from mathematical slop. Enter the “not by AI” badge.
Basically, it’s exactly what it sounds like. A sleek, modern badge that you slap on your artwork to tell people that you did this, not an AI. There are pre-baked versions for writers (“written by human”), visual artists (“painted by human”), and musicians (“produced by human”). The idea is that these badges would help people identify human-generated content and steer away from AI content if they’re trying to avoid it.
It’s not just intended to be added to individual artworks. Websites that have “at least 90%” of content created by humans are invited to host the badge, along with apps, too. This directive reveals an immediate flaw—the badge would easily confuse someone if they read the 10% of content by AI on a site wearing the badge. There’s also nothing stopping people from slapping the badge on AI-generated content and simply lying to people.
You might take a more cynical view if you dig deeper, though. The company is charging for various things, such as a monthly fee for businesses that want to display the badges.
We’ve talked about this before when we asked a simple question—how do you convince people your artwork was made by a human? We’re not sure we’ve yet found the answer, but this badge program is at least trying to do something about the issue. Share your human thoughts in the comments below.
Works great – until someone just puts it on AI generated content.
The reason the ‘kosher’ or ‘halal’ or ‘UL’ tags work is because there’s a group that investigates and enforces the use of those marks – up to and including using violence from the government to enforce it.
Yep. If the fee went to someone who actively confirmed the content came from humans and could revoke misused badges, the fee might make sense. If they let just anyone download the badge, why bother?
I don’t know about kosher or halal, but I’m pretty sure that 50% of all the extension cords and chargers on Amazon that claim to be UL-approved are not.
But fundamentally, why this couldn’t be different? If it gains popularity and they get funding, and they uncover misuse, they can certainly send cease-and-desist letters or file lawsuits.
Yeah all the “solutions” to this “problem” have been pretty universally pathetic
“We’re not sure we’ve yet found the answer, but this badge program is at least trying to do something about the issue. ”
They’re really not. They’re trying to make a quick buck from gullible people.
Today on scam gets pushed by hackaday
I was hoping there would be some cryptographic trick that actually gave this some teeth. An image that you pay a monthly fee to have copyright to, with no enforcement on what it goes on, is meaningless.
This type of product deserves to be critiqued, not covered. There’s a difference. Thick critically people.
Umm… that’s what we’re here doing?
We’re certainly not endorsing it.
Not sure a badge of questionable worth is worth $13/mo
“May contain AI”
snake oil….
I don’t care. If something is good, but came from AI, I’ll take it. If it’s poor, but came from a human, I’ll reject it.
The problem is that all the current AI just vomit out what they have taken from human generated content with a tiny bit of randomness, a tiny bit from here a tiny bit from there shuffled about a lot.
When almost all content is AI generated then you eventually suffer from garbage in garbage out. There is no fresh content entering the system, it is just recycling everything that has existed before, shifting the pieces about a lot. And at that stage human will cost too much compared to “AI” regurgitated content.
The current set of neural networks are not smart, and have no actual intelligence. How they appears to be intelligent to some people is because they has absorbed more content than a single human could access in one lifetime, but at a truly astounding energy cost to train the individual networks.
Correct, but if it vomits out something good, I’ll take it. My opinion is the final arbiter, not whether or not it was produced by a human or AI. At the moment I’m mostly unimpressed with the vast sea of garbage on YouTube, for example, that is human generated. Adding more stuff for me to ignore doesn’t matter.
yep, it’s not like if humans never used any prior art…
Wow, tell me you don’t understand generative AI at all without saying “I don’t understand generative AI at all.”
It’s not that he’s not understanding AI, it’s that he’s not caring. Much like people who don’t care about factor farmed chicken .v. ethical farming. They just want chicken. This guy just wants the content and doesn’t care about how it was produced.
There’s a tragedy of the commons in effect here, but it’s not an unreasonable position to take if you’re not an artist
The other problem is that the vast majority of human creators do the same thing.
Bad artists are not at fault for being out of work but if they get scammed by these stickers, its 100% on them.
So, what does the $13/mo buy you? A license to youse their specific image? I’ll just ask an AI to generate a similar looking badge.
They don’t even actually verify the content and use the honor system? This is about as useful as when twitter sold the blue check ark for a few dollars.
Did they pay you to write this article?
I’ve never commented on a Hackaday article before, despite being subscribed for years.
I will admit that the tone of this one was what brought me here. I understand that strong emotions are attached to this issue, but dismissing all AI generated output as dreck and further inflaming the debate with this profiteering badge serves no one except the person making money off other people’s anger and fear.
Careful, considerate debate is needed on this issue. Reducing it to moralizing and character assassination just further entrenches both sides. The issue is filled with far too many subtleties, and has far too much potential for both harm and benefit to relegate it to a battlefield with only two sides, both insisting they are the good guys.
Absolutely agree with this statement. There’s no point polarizing such a complex topic into “right” and “wrong”. This needs to be approached with reason!
What if reason leads you to “right” and “wrong”? Nuance is not necessarily a virtue.
Ok, so I give them my email address and they send me a bunch of stickers (images) for my own personal use. Now I can go slap these on anything and everything I like right?
Ok guv’nor, I promise, cross me heart and all that, ONLY on stuff I make meeself (I can still use ChatGPT as me spel chekker rite?)
Oh, while I got your attention, can I interest anyone in any 4 speed walking sticks? Chocolate teapots? Ooo got a special on perpetual motion machines this week too… all completely genuine, honest!
Guess they used AI to produce these badges …
Also how is writing not producing? What if my music contains lyrics?
Actually, I think we should do it the other way. All AI generated images of people (and animals, I suppose) should have a capital ‘H’ on their forehead.
Arnold Rimmer may have something to say about that…
Also paying $12.99 per month for a badge is insane
For that price, I could get 1.5 blue checkmarks!
For the same price a person can get unlimited AI and a page at Graydient.ai though
which will catch on
Hey, nice human-generated art ya got here. It’d be a shame if somebody accused you of usin AI. Me and duh boys can protect you from dat. As long as you keep makin yer monthly payments.
Yep. 100%. This is the TL;DR
You should pull this article. At $13/month for just some jpegs and absolutely nothing else, this is an obvious scam.
Honestly pretty disappointed to see this level of dreck on HaD
> $13/month for just some jpegs and absolutely nothing else, this is an obvious scam.
Hey, plenty of NFTs got sold just like that!
This article just does not sit right with me…
Paying monthly for a badge someone with slightly above average inkscape skills could put together in an hour, with no other benefits? This seems like someone is trying to make a quick buck off of artists.
We don’t need stupid stickers. We need a class action lawsuit on behalf of ALL owners of intellectual property appearing on the internet. This would be the greatest legal case of its kind in history.
And what is my argument? Simple.
I have a website. You can visit and enjoy the content for free. What you can’t (legally) do is download it, repackage it, and sell it as part of YOUR product… not without my expressed permission. This is the case with many (most?) such sites.
Repacking my data to a different data format doesn’t get you off the hook. My IP is the -content- of my files, not the manner in which it has been encoded into a data structure.
And that’s the point: If the ChatGTP folks scraped the internet and used copyrighted text, images, and other content to train their AI, then parts of all of those properties have now become encoded into ChatGPT’s data structures. And that’s theft.
This is no different than a hypothetical case where a thief scrapes a jpg from my site, converts it to a png, and then attempts to sell it as his own.
If the ChatGTP folks had commissioned 10,000 artists to paint/draw 10,000 images to use as a training set (and paid these people for their time) , fine. Operating in gray areas of internet law to steal from thousands (millions?) of content creatures is NOT fine.
I can still remember a grandmother being criminally prosecuted because her 13 year old grandson had (gasp) downloaded unauthorized MP3s to his player using her internet connection…. but here, industrial-scale theft of the whole internet’s IP is… “progress?”
Let’s start with a subpoena for a list of all URLs which the ChatGTP crawler visited. If they can’t provide that, then follow with a Federal court order to erase their trained structures and start over with material to which they actually have demonstable right.
You are… Massively uninformed about how both copyright law and LLMs/”AI” work.
Seconding this. Normally I’d break down the argument and counter the points, but at this point it seems clear that this user did not research the currently ongoing lawsuits, previous lawsuits, and government information provided on this topic, nor do they understand how A.I works. The boundaries of generative A.I are currently being developed by ongoing lawsuits. Their version seems like a daydream.
Probably should also sue those thieves that looked at your art and encoded it into their brains. That’s theft.
This might actually backfire spectacularly, considering that nobody still knows whether long term improvements in AI will need fresh content and actually have to avoid AI generated content like the plague to avoid over correlation in the training dataset.
By doing this, you actually are signaling that this information is going to improve your AI model, and are helping the issue these badges should be fighting against.
And honestly, asking 13$ for something that should be freely distributed to all of this following the same cause is pretty scummy, like others have said.
> nobody still knows whether long term improvements in AI will need fresh content and actually have to avoid AI generated content like the plague
Search for “Model Autophagy Disorder (MAD)”. AI-generated images used to a train new inbred AI creates an unwanted positive feedback loop where the results get worse. More convergence towards a lower number of possible images that can be output. And the quality of the images generated drops due to the positive feedback loops. All future AI will need to filter out all AI generated content!
If a writer wrote the ungrammatical nonsense “written by human” on their work I’d automatically disregard their content anyway.
>90% human and not AI!? How about 100% not AI.
.
Also all the other stuff people said. This is horseshit.
its stupid crazy greedy $#1tty western guys and poor indians (citizens of India), as well smartass stealer “startups” that compromised notion for AI. Ultrastupid westerns ,mostly from USA are educated to see money in everything ,to compromise it quickly . AI is a good and useful thing.But not when promoted and sold buy shitty US (mostly,but not only ) guys. They compromised many other technologies in a grade that when normal companies hears of it,they run quickly and says No,no,please . As such” startups should be strongly discouraged ,AI works ONLY !in very big corporations (GAFAM) and scientific institutions . Anybody else should be strictly disallowed to work with AI,even more garage “startups” that a liers,scammers and stealers of money .
>AI works ONLY !in very big corporations (GAFAM) and scientific institutions . Anybody else should be strictly disallowed to work with AI
Yeah, no thanks. If anything it should be the other way around. Also, you should probably ask ChatGPT for some English lessons.
I’ve also read hackaday for years and never commented before. I would have liked to see a lot more skepticism rather than treating this as an earnest attempt at a solution. With no mechanism for enforcement this is an obvious scam.
Covered in the post, although perhaps not as strongly as you’d like?
1) 90% human.
2) They’re trying to get money out of you for a JPEG.
This is cope, you guys wrote an ad.