I recently touched on the role that A.I plays in contributing to my Dead Kulture Theory. It’s a huge topic with many discussions surrounding it. I knew I needed a follow up to expand further on the pros and cons, as well as my experience so far with A.i.
First off, I want to state that I am a huge fan of A.i, the current and possible capabilities is very exciting. A.i is a powerful tool that artists can use to level up their game in ways they never thought of before. That being said, there are some concerning things happening around the software.
To me, there are a few parallels between this new tech and social media. Social Media had immense benefits for creatives and brands. So much so that many of us ignored the negatives until very recently. I am not trying to do that with A.i. I'm trying to be very conscious of everything I can be.
Not everyone shares my interest in the negative though. Many won’t even acknowledge there is any. Which is insane. We’re going to cover some of those negatives A.i has had and will have moving forward in the art world.
Problem 1: Unethical Datasets
One of the most important negatives we run into. Is when we examine the datasets that these software’s are learning from. After some reading over at exposing.ai a clear pattern emerges. User’s work, marked no commercial use and attribution required. Are ending up in these pools without consent. No attribution and most the time no knowledge, certainly no compensation for taking part in the development of a product valued at billions.
To make matters worse, you are able to train these A.i’s off one specific artist. For artists like, Greg Rutkowski and Hollie Mengert this has become a real problem.
You can read about Hollie’s Story and hear her thoughts on all of it, in a nice coverage put together by Andy Baio.
As for Greg’s story and his thoughts, you can read a great piece by Melissa Heikkiläarchive.
Now, you might not care about those issues those artists are facing. You might not care if your image is used to make art.
That being said, I would think you might (or should) care if the government uses these datasets for surveillance tech. Which they are. What about when certain corporations push beyond cute profile pics with this tech? We’ve already seen so much unethical use of data when it comes to social media, more about this later.
Check to see if Diffusion or Midjourney used your art:
https://haveibeentrained.com/
The argument against everything I just said: varies from what I can tell, from some people believing anything on the internet falls under public domain(not true), to some people believing its transformative(currently up for debate/depends/ethics) and you can’t copyright a style(which of course is true).
People also note that this is what humans do, get influenced/inspired from other artists work and it can show in their output.
To that I would say: It’s not the same thing, not to me at least, again its mathematical, no soul behind it. The fact these datasets are filled with people’s content who did not consent and get no attribution in which then get utilized by 3rd parties for however they see fit is dangerous, and to me seems unlawful. Also, I can create without an algorithm/other peoples work, this software is dependent on it.
Problem 2: Bad Actors/Bad Faith
The next negative you run into is a wide array of bad actors abusing the tech. It can be the dataset creators, or the 3rd parties utilizing said data to create products. Many times these are the same people who have created a new entity to handle the dataset as a form of risk management from their main company.
Then you have the everyday users of the tech of will find all sorts of creative ways to abuse the tech. This could be an argument to how the tech is fueling creativity, not killing it. I think both can be true.
Of course this isn’t the tech’s fault. Though its important to discuss to promote finding any possible solutions. Anything to help curb the negatives or raise awareness, so we can better enjoy the positives.
Manipulation
There’s all sorts of situations where this tech could be used in bad faith. We have seen as much with the likes of social media, which is why I mentioned parallels. If you don’t know what I’m talking about check into the Cambridge Analytical Scandal. As well as watch the Lincoln Project documentary. It shows the astounding capabilities of mass manipulation and fraud through social media.
Harassment
Malicious parties will use images, videos, and sound from these A.i’s. That they otherwise wouldn’t have access to. To mislead viewers many ways. Character assassination and scams to name a few.
“Models fine-tuned on images of exes, co-workers, popular targets of online harassment campaigns. Combining those with any of the emerging NSFW models trained on large corpuses of porn is a disturbing inevitability.” — Andy Baio
Copyright Issues
Then you have IP and copyright issues that will arise. People selling copyrighted material directly on products. Sometimes breaking two copyrights depending on the A.I they use.
Dishonesty
You will also have people who are not upfront about using an A.i, and claim they created it themselves. To be deceitful to a client about anything. Let alone something that would affect their decision making process. Is a horrible look, bad practice, and should be shunned.
This also takes away from what pro A.i-ers use as support in their argument. Saying that A.i will add value to human content. This won’t be the case if everyone is lying about what is and isn’t A.i generated.
-
The argument against everything I just said: Is that all this stuff already happens without A.i and will continue to happen, with or without it. You can not blame the bad actors behaviors on the software itself.
My reply to that is: I agree, but again its important to be conscious of all this and form discussions to perhaps make some of those things better in this space. Can we not talk about it? I am in no ways suggesting… anything really, certainly not that we should close the book on A.i simply because of bad actors. If we did that for everything we would have nothing.
Problem 3: Killing Kulture.
The next negative is the one that hits me the hardest. A.i in the art world is contributing heavily to the Dead Kulture Theory. In other words, it is in ways stripping the soul out of the creative process as I talked about in the theory.
I already see it all over my feeds. The gurus telling everyone to grab the highest performing video on YouTube. feed it into A.i to generate a script. Feed that script into another A.i for some sound, then another A.i for a video. Then once you have that, put it through another A.i blender for shits and giggles. Other than the copy & paste, when does the human become involved in this process? Is that what artists will be reduced to, copy & paste clickers?
That’s the creation side. Think of the consumption side. Accounts will be/are connected to a.i to like, share, and comment on content. So now you have an ecosystem that’s largely artificial. Bots liking and commenting on posts by bots, and the humans will get lost in the spam comments.
https://twitter.com/i/status/1592652469691363328
ChatGPT is here!
— This video talks about the prospect of having A.i write your SEO pages. Do you think an A.i can capture the soul of your brand?
Enable 3rd party cookies or use another browser
— This video talks about the prospect of using A.i to write and send cold emails. Sounds like a nightmare. Its already so apparent when these cold emails are copy and paste templates, no personalization. Its basically spam. The A.i writers I’ve seen are not that good.
People are raving about ChatGPT, its cool and for sure the best I’ve seen. To me it almost seems like its just a google interface you know? Instead of searching and clicking around for an answer, feels like this A.i will just tell you what it thinks your looking for in a paraphrased way.
When google snippets dropped, I thought it was good for users but bad for the content creators. Sure there page is featured on top of the results page, but if my answer is sufficiently answered in that snippet. I no longer need to visit their page. Less page hits can mean less ad revenue and less sales.
This makes that type of thing much worse. At least with google snippets there’s still a link to the page. This A.i is using datasets, that scrape all of those articles, to put something new together. Again with no proper citations or anything. You can argue its transformative, but it still sucks for these writers. Whose work is being used without consent, attribution, or compensation. In a way that can both give and take attention from them totally outside of their control.
People say this will add value to real human work. Which would be true if every A.i artist was upfront about their work being product of A.i software. People are already being dishonest about the use of these softwares.
The quality of the text generated by GPT-3 is so high that it can be difficult to determine whether or not it was written by a human, which has both benefits and risks.[4] Thirty-one OpenAI researchers and engineers presented the original May 28, 2020 paper introducing GPT-3. In their paper, they warned of GPT-3’s potential dangers and called for research to mitigate risk.[1]: 34 David Chalmers, an Australian philosopher, described GPT-3 as “one of the most interesting and important AI systems ever produced.”[5] An April 2022 review in The New York Times described GPT-3’s capabilities as being able to write original prose with fluency equivalent to that of a human.[6] -https://en.wikipedia.org/wiki/GPT-3
Why GPT-3 is the best and worst of AI right now
We were taught to cite every single source we used in the creation of our writings. We were taught this vigorously, I remember them making such a huge deal about it, because it is.
New Wave of Kreativity
People are talking about how this is going to usher in a new wave of creativity. Do you know why that’s true? Because of my Dead Kulture Theory. Everyone’s already copying each other without A.i. The main wave seems bent on going all the way. We already see people copying each others prompts. Which you would think wouldn’t be a problem but it is for some. Which I find leads to a hypocritical situation(see meme below). Now even when they copy each other’s prompts, they will get a different result.
There will be all sorts of things happening with it. Without injecting ones soul into a piece, and having to rely on an algorithm and others art. It seems like a lot of bad news for the creative process of human expression in my humble opinion.
“We’re applying a capitalist streamlining logic to literally our emotions and our souls” — Bo Burnham
I also don’t like the attitude that some users have about it. You would think they would at least be grateful or sympathetic. Especially after they get upset when someone steals their prompts.
Just like the guy in the video above, “this is why im beating you guys”… your only “beating us” because you have to leverage someone else’s content, you are technically a conman and no true creative.
Will this make us Dumb, less skilled, and more reliant on machines?
Students are using these tools to write their homework assignments. I can’t blame them, I would have used the tools if I had them when I was in school. I still wonder what long term effect that has though, if any. Best we had was like cliff notes. We still had to ingest the info and write a dam paper, we didn’t have no machine to spit out a paper in under a minute.
It’s a testament to how every single aspect of our lives is becoming more and more digital. Our lives are becoming increasingly dependent on technology. What are we going to do when the 4th industrial revolution finally happens? When robots are doing everything for us? Talk to each other? Fat Chance, our A.i avatars will already be doing that in the metaverse.
Solutions
A.i should only be trained off public domain images for start. It be nice if you couldn’t train these programs on specific artists. Something that Stable Diffusion 2 has taken away, and has prompters upset. Content Authenticity Org. is working on some features to track authenticity of content thru attribution solutions. People could also use it ethically so it wouldn’t be an issue, but that’s asking to much of people.
To me, using A.i ethically means;
To use it as a base plate with heavy modification, reference/study, or inspiration.
Being upfront about the A.i use and include any specific artist prompt.
Never using an A.i result with no edits in commercial settings.
That’s how you keep it in line with the whole argument of “this is what humans already do”.
Clarifications and Conclusions
Again, I am a fan of technology and excited to see where this sector goes. I just see a nightmare that comes along with that effecting so many different things.
Just like with anything, what I speak of does not apply to everyone.
When you discuss any of these negatives. Die hard supporters of A.i get defensive but project that defensiveness on to you. Accusing you of being a purist fighting progress, or an artist that’s threatened by A.i. Which is an absurd assumption, don’t assume every artist is afraid of being replaced. True artists are confident in their abilities to create and adapt. Many are like me who embrace the technology while being conscious of the downfalls that come with it.
Which brings me to my next point. No matter what we think about it, no matter what we do about the negatives. A.i art is here to stay and will no doubt change the creative landscape moving forward. Better to try and embrace it and use it as a tool to at least learn the fundamentals to stay up to date with current trends.
As I said in the Dead Kulture Theory, A.i art will not kill creativity, or the brush stroke. There will always be artists expressing themselves through different mediums, formats, and tools. There is a lot of noise and hysteria but it is not the end of the world(even though I love acting like it).
There are still plenty of artists with soul out there capturing. Expressing the human experience in a unique way that promotes their individual self.
Think of it this way. A.i never produces a result anywhere close to what’s in your head. Meaning, what’s in your head, is still valuable.
A.i is not sentient(as far as we know lol). Yes these software's work off algorithms and other artwork, but it still needs a human to direct it on both ends.
What I guess I’m saying is, it’s on all of us to be more conscious about these things and just be better in the spaces in which we operate. Use A.i as a tool to level up your creative process, not as a replacement.
ai is moving so fast, I don't think people realize where its going
— Evan BealeUntil the programs themselves can change fundamentally, it won’t matter much outside of hobbyist use anyway. These programs are fun to mess around with, but you can’t use the end results in a professional or commercial sense the majority of the time anyway, so their practical use is very limited. If and when they are able to redevelop the tools to work without essentially stealing from already existing — and often copyright protected — inputs, then we could indeed be looking at the future of asset creation. But until then, it’s just a fun toy and not much else.
— Ninja Tonyrespectfully, this seems to be a limited view, im looking at big picture and the rate at which these programs are advancing and its insane… i think majority of people will be surprised where this sector goes..im surprised every week lmao, it may not be THE future of creation, but it will certainly play its part. its not a fun toy, well it is lol, but i would describe it as an incredibly powerful tool that more artists will utilize more and more in the future, we are seeing it in real time more and more everyday in the creative space. I would say the practicality of these softwares are limited to your imagination, they are great for artists to use as base plates, references or inspiration that can be used in commercial settings. Artists are using a.i all sorts of ways to up their game. Unfortunately there is a great deal of people out there creating art for others using these software’s and not changing much, which is a major problem for the same reason a.i, as is, will never replace artists. Its not sentient, all A.i is at this stage is data moshing human generated content, in a sense. Whether or not we ever create true artificial intelligence that’s sentient is a whole other huge ass topic lol. BUT no doubt a.i will continue to improve exponentially as we are seeing in real time and it will no doubt find its mainstream place in the creative space being used in positive awesome ways and bad negative illegal ways by a growing number of “creatives” going forward no matter how we all feel about it
— Evan Beale
It seems like every year, technology advances faster and faster. We are in for a wild ride, I’m not sure how many people actually think about it, but we are going to see some wild shit in our life time. I would advise to learn as much as you can about it all, the good, the bad, and the ugly. Don’t dismiss it, embrace it. Nothing can kill true artistry unless we let it.