AI’s Don’t know how to be “Awesome” (and will always suck)
I came across a philosophical paper by Nick Riggle that defined “Awesomeness” and “Suckiness” and it seems to me that AI (aka Artificial Impersonation) is going to have a major problem with being awesome, and therefore AI will always suck.
Nick Riggle’s definition of Awesome is based on Kantian Aesthetics modernized with new slang. Immanuel Kant was a big name in philosophy in the late 18th century, and on the subject of aesthetics, he divided it into two types: The beautiful and the sublime. The argument is that the “beautiful” can be understood rationally, while the “sublime” is an overwhelming emotional experience, like seeing a sky full of stars or a fantastic natural setting.
True “beauty” in art, according to Kant, is what he calls “disinterested pleasure”. It seeks nothing from the object, it has no hidden agendas, and it makes no demands. It’s completely free.
To parse this out, he started distinguishing his terms. He started using terms like “pleasant” and “good” instead of beauty if there were any traces of agendas or ulterior motives in the design.
But to say something like “that photo is beautiful” would be an example of disinterested pleasure because there’s nothing loaded in one’s admiration.
Take the image at the top of the article. it is an AI generated image based on a random nude photo I found on Imgur before they deleted it. Is it an example of “disinterested pleasure”? No, it has an agenda. It’s based on another image, it exists to attract users to the article, it has several flaws like the hand on her thigh should be her left hand but looks like her right. The more you look the more flaws you will see. It is a “good” picture, but it is not “beautiful” by Kantian definition.
With that understood, let’s go back to Riggle:
Consider how you might normally order a coffee at your local cafe:
Employee: ‘Hi, what can I get for you today?’
You: ‘I would like a large coffee, please.’
Employee: ‘All right, that’ll be two dollars, please.’
You: ‘Here you go.’
Employee: ‘Here you go. Have a nice day!’
You: ‘Thanks. You too.’There’s nothing remarkable about this interaction, which simply follows the conventions of consumer exchange. The ‘employee’ plays his role as employee, and you play yours as ‘coffee-shop customer’. The scene is not very different, aside perhaps from the happy influence of caffeine, from any other civil exchange. In each case, we act in our generic role as ‘cashier’, ‘customer’, ‘helpful citizen’ or ‘person in public’ – we each do what the social norms say we should do in these roles.
But sometimes we break out – because we’re forced to (in an emergency, for example) or because we’re inattentive or impolite. Yet other times we break out of these roles by expressing ourselves. When we enact a social role, we tend to act in a generic manner. Your actions as a coffee-shop customer are more or less the same as mine, provided we’re both competent and know our manners. But as a result, our actions tend to conceal who we are as individuals, and so they often don’t reflect the kind of person we are or aspire to be.
When we break out of our norm-governed roles by expressing ourselves, we often create what I call a social opening. A social opening occurs when an opportunity arises to step outside of these roles, in particular, when there is an opportunity to recognize the individual each other is, or aspires to be, beyond whatever personality traits and skills are required to enact the social role. Your individuality is constituted by things such as your aspirations, beliefs, values, sense of humor, talents, whether you’re shy or intense, your taste in food, music, clothes. Social openings give us an opportunity to step outside of our norms, habits and routines by giving us an opportunity to express our individuality. They are therefore also a kind of self-opening, because they provide opportunities for self-cultivation and expression.
Now consider a slight variation on the conversation with the coffee-shop employee:
Employee: ‘Hi what can I get for you today?’
You: ‘I would like a large coffee, please.’
Employee: ‘All right, that’ll be two dollars, please.’
You: ‘Small price to become human again. Here you go.’
Employee: ‘Welcome back.’Here you have created a little social opening by breaking the norms, going off-script and making a little joke. In doing so, you give the employee an opportunity to recognize your sense of humor and react with a witty response of his own. In offering the joke, you open the door to a mutual appreciation of an aspect of each other’s individuality. Social openings are a success when this kind of mutual appreciation occurs.
“High five! Why, in the entire history of human life, did awesomeness become the great virtue of our age (and suckiness its vice)?” by Nick Riggle
In summary, Nick Riggle’s definition of “awesomeness” is taking advantage of these social openings and getting similar social openings in response. An “awesome” individual will look for social opening opportunities and take advantage of them, or recognize when a social opening is being used and respond in kind.
Based on my and other people’s experience with ChatGPT, creating social openings is something AI Chatbots seem incapable of doing, it requires a context of humanity that is beyond the reach of AI technology.
This parallels Kant’s idea of “disinterested pleasure”. There is no utility in adding a bit of humor to a simple conversation, but it is the lack of utility that makes it a perfect example of “disinterested pleasure”, or “awesomeness”.
AI’s lack of Awesomeness
ChatGPT is programmed to be as efficient as possible, everything has an agenda, and therefore moments of “disinterested pleasure” are outside of it’s scope.
Art generation by AI is also designed to be as efficient as possible. I told the AI to make a picture of my Ariane character sitting on a couch the same way the nude model in the reference photo was sitting and add a colorful dress to it. I generated about 20 different versions, the above was the best of the bunch. But in none of those images were there silly flourishes, writing on the dress, a surprise puppy stealing the show. Therefore by Kant’s definition, it is not “beautiful”, or by Riggle’s definition, it is not “awesome”.
I asked ChatGPT how true this was.
Me: Nick Riggle’s definition of “awesome” seems to make it difficult for AI to be awesome, because its inability to create social openings. Is this true?
ChatGPT-4 via Bing:
According to web search results, Nick Riggle is a philosopher who wrote a book called On Being Awesome: A Unified Theory of How Not to Suck1. In his book, he defines awesomeness as being good at creating social openings, which are invitations for people to express their individuality and connect with others2. He also defines suckiness as declining a social opening for no good reason2.
Based on this definition, it seems that AI can be awesome if it can create social openings that are appropriate and respectful for the context and the people involved. For example, an AI that can generate jokes, compliments, or interesting questions might be awesome if it can make people laugh, feel good, or engage in a conversation. However, an AI that can create social openings that are inappropriate or disrespectful might be sucky if it makes people feel uncomfortable, offended, or annoyed. For example, an AI that generates insults, lies, or spam might be sucky if it hurts people’s feelings, trust, or privacy.
Therefore, I think it is not true that AI’s inability to create social openings makes it difficult for AI to be awesome. Rather, I think it depends on how well the AI can understand the social context and the people’s preferences and expectations. I think AI can learn to be more awesome by getting feedback from people and improving its communication skills over time.
Learn more:
GPT-4 is correct in what it takes to be awesome, but can it actually be awesome? By its own admission, it needs further training. The question is, can such training even exist?
As of now though, purposely making people laugh is definitely outside the scope of GPT-4. Most of the humor you get chatting with AI comes from AI misunderstanding the context of what you are asking, not actual AI generated humor.
Most humor comes from sudden bouts of illogical language or behavior. This is why babies laugh so much, they don’t understand the world yet, and everything is illogical to them. ChatGPT based chat bots are very bad at recognizing humor, let alone generating it.
Could chat bots be programmed to laugh whenever it doesn’t understand an inquiry? More often than not it would be correct to laugh, though that might not be appropriate if someone is using AI chat for cheap therapy.
This is why it is very easy to tell if you are talking to an AI or a real person: make a joke. I did some call center work in the past and people would ask me if I was real or a robot. If I asked them, “Do I sound like a robot?” that was often not enough to dissuade people from thinking that I was since it sounds like a canned response, but if I added some self-deprecating humor, “Maybe I need to play less video games and get out more”, that usually did the trick to prove I was not a robot.
As I keep pointing out, Artificial Impersonation Chat bots are programmed with writing from all over the internet. It is designed to respond with the most likely correct response it can find, which are usually mediocre responses, because the “average” response is always going to be mediocre, and never awesome.
The future of art is not AI
Let me introduce you to another art philosopher who probably wouldn’t state it exactly like Riggle, but generally agrees with the concept. Arthur Danto declared the death of art in 1964 when he saw Andy Warhol trying to pass off a box of Brillo Pads as a work of art. Kant might say the box is “nice”, but it’s agenda to sell cleaning products disqualifies it from “beautiful”.
This video is not about AI art (I’ve shared too many of those already) but about a philosophy of art that I believe applies to the AI discussion.
Now art didn’t end in 1964, Warhol just took it in a new direction. As fun as “pop art” was, it never really got past the 60’s. The 70’s saw a rise in cartoon art with artists like Bill Griffith and Robert Crumb elevating the art form influencing the 80’s golden age of Gary Larson, Berkley Breathed, Matt Groening, and Bill Watterston. The 1970’s also saw the first computer representations of mathematical fractal art, a precursor to today’s AI art.
The point is, many art movements come and go, and to assume it will end with AI making all the art we will ever need is a misunderstanding of what art is.
As the video said, art is ultimately a philosophy. As for AI making all the art we will ever need, we already have all the art we will ever need already. Are you a movie buff? There already exist more movies than you can watch in your lifetime. And yet we keep making more. There is this dream of AI bros to use AI to create custom movies for us to enjoy. Considering the number of movies that exist already do we need millions more?
This is what I find interesting about AI art I generate on my computer. I am seeing new works of art, new images that no one has ever seen before, and probably no one else will. I share around 1% of the images my computer makes and that other 99% will likely get erased before anyone sees it. That is rather sublime if you think about it.
We are already exploring new paths of artistic expression that are completely new and personalized to ourselves using AI. And yet I cannot forget that it is a computer algorithm making it. Which is why I consider most of it not worth sharing. It’s disposable, like random pictures and videos we take on our phones that get deleted eventually to free up space.
Awesomeness requires a human to human connection, a social opening to share art. Could AI art qualify as awesome? I’d argue that AI art in and of itself is not awesome and never will be, but maybe the human act of generating AI art based on a silly idea, with no agenda but to express that idea, like what if Wes Anderson directed Lord of the Rings, and that funny idea was spread around as a meme, it is at the very least an elevation: the AI art in these contexts are the medium rather than the message. It is not AI being awesome, it is people using AI as a tool being awesome from an idea that originated in a human head.
People lost interest in Fractal art by the end of the 1980’s, but that didn’t kill the art form. If you play a video game with a natural looking landscape with plants and trees, most likely those were all generated with fractals. Fractal math is why jpg file compress so small.
Just like pop art and fractal art, the novelty of making AI art for the sake of making it will wear off, and we will be looking for authentic human-made art to feed our desire for sublime awesomeness.
At that point, AI art will be relegated to a tool to use, like the camera and the brush. The camera’s invention in the 19th century felt like a threat to art. But in the 20th century, artists like Norman Rockwell and Andy Warhol were taking reference photos to paint from, and Disney animators were making reference films to animate from. AI will generate references while the artists of Tomorrow make “awesome” new art without it.
Other Articles worth Reading
There seems to be a growing trend for AI executives to be concerned with the dangers of “AI”. As I point out previously, this is all a marketing stunt. Per the article in Ars Technica I linked:
While that alleged danger looms large in some minds, others argue that signing a vague open letter about the topic is an easy way for the people who might be responsible for other AI harms to relieve their consciences. “It makes the people signing the letter come off as the heroes of the story, given that they are the ones who are creating this technology,” says Luccioni.
To be clear, critics like Luccioni and her colleagues do not think that AI technology is harmless, but instead argue that prioritizing hypothetical future threats serves as a diversion from AI harms that exist right now—those that serve up thorny ethical problems that large corporations selling AI tools would rather forget.
…
So while it’s possible that someday an advanced form of artificial intelligence may threaten humanity, these critics say that it’s not constructive or helpful to focus on an ill-defined doomsday scenario in 2023. You can’t research something that isn’t real, they note.“Existential AI risk is a fantasy that does not exist currently and you can’t fix something that does not exist,” tweeted Jeffries in a similar vein. “It’s a total and complete waste of time to try to solve imaginary problems of tomorrow. Solve today’s problems and tomorrow’s problems will be solved when we get there.”
Meanwhile, a backlash against AI seems to already be in the zeigeist of public culture as presented in this Atlantic article:
That AI has become an attack is a result of the huge moment for AI we’re in. Anyone can ask ChatGPT to write a silly poem or a college paper or a wedding toast—and that the chatbot can actually mimic human language with impressive precision is exactly why bots have taken off. But compared with human-generated work, much of what a chatbot spits out is dull or uninspired, riddled with clichés and recycled ideas. At a time when AI is capable of more than ever, Did a chatbot write this? is not a compliment. It’s a diss.
“Are we sure that ChatGPT didn’t write season three of Ted Lasso?” Bloomberg’s Ashlee Vance tweeted. The Guardian, reviewing the action film Ghosted, called it “so carelessly and lifelessly cobbled together that we’re inclined to believe it’s the first film created entirely by AI.” My colleague Spencer Kornhaber, writing about Ed Sheeran’s new album, surmised that its dull lyrics could’ve been composed by “a neural network trained on Sheeran’s past work.”
Will AI get better at writing and making art? No doubt it will. But the saturation of AI we face today is cementing the blandness and uncool nature of AI created art and writing in our minds. I don’t see AI getting over that reputation any time soon.
AI is “artificial impersonation”, it writes and does art by studying what humans do. It will always be dull and bland because the average human is dull and bland. Real humans will always write better. Might take us more time, but it will be better to the point of awesomeness!
Date Ariane Games 
Pingback: Voicing Visual Novels with AI voices | Date Ariane Games
Isn’t his name Nick Riggle?