The Uncanny Valley of Good & Evil
How to make a frenemy of artificial intelligence in the creative landscape
This month I have been pondering the shift that ChatGPT and generative bots will bring and is now bringing to industries that include writing, creativity, images, and ideation. Podcasts, articles, town halls – everywhere you look, there is a ton of chatter about artificial intelligence and the sway of AI on creative industries. This year, AI took a huge leap forward. The capabilities are accessible, open, and updating every day. The term “uncanny valley”, coined by Japanese scientist Masahiro Mori, is used to describe the feeling of connection with a human-like bot but then knowing suddenly that you are not speaking to a human entity, but to a robot. A veil is lifted as your brain remembers this is not a real person. Because we live in a world of language, interacting with words makes us feel that we are interacting with someone. When in reality — are we? It’s becoming clear this valley is a place we may now always inhabit.
I’ve been processing my perspective on all this because I feel — I must? The Writers Guild of America - poised for strike by the way - is suggesting writing into their new contracts that AI be able to be used. The writer would still get the credit, and AI would not be acknowledged in the writing credits. It seems that we need an ethical approach where there is no ethic, to begin sorting through our knowledge of something that will likely change everything about creative industries.
For one thing, I encourage folks who have not tried these free programs to do so. You can work with the OpenAI ChatGPT to edit, write copy, make lists, create writing prompts (see below for my super simple ChatGPT exchange), or help you organize your thoughts. The most recent update can now connect to the internet and browse in real time—meaning it can do a lot of very fast research for you. For code writers and web developers, the bot can actually create a website, create images, and build programs from simple ideas. So, for the creative community, people who make a living doing many of these tasks, many are asking: can AI do what we do?
As The Daily recently said about deepfake songs and audio, “it’s fun, but it’s kind of a novelty.” That’s true. But it’s also constantly evolving. It will be be new over and over. It will not stop, the way all computer technology is able and has been able to do in constant iteration for the past 35 years or so.
The first thing I have landed on is: There’s no stopping this thing. It can turn words into code into interfaces. It can take our ideas and make them better. Which basically means, it can put imagination into creation, taking simple words and turning them into art in seconds. It can bring our thoughts to life, instantly.
I share this thread that I saw a few weeks back on Twitter, which was really the lightbulb moment for me. A user of the most recent update to ChatGPT-4 gave the robot a simple instruction: Starting with $100, make me as much money you can, as fast as possible. It proceeded to write the code for a website, create a business plan and idea, and even publish the site with images created from Dall-e (which, by the way, if you haven’t used Dall-E or MidJourney [in my opinion MidJourney is much better], I recommend taking a spin!).
Here is the website it created from his prompt, complete with AI generated content, images, and basic business plan:
https://www.greengadgetguru.com/
You can read through the thread here:
https://twitter.com/jacksonfall/status/1636107218859745286?s=20
This isn’t like…what I want to do. But it is true that as much as I want to produce media in a beautiful and impactful way, I also want to make things in an efficient way. I am also seeking sustainability for my family life and home. I want to make things that are good. But like any worker in the inescapable capitalist flow that we exist in currently, I also want to cut out the parts of my work that don’t come as easily or make me feel like I’m losing time. I want to delegate. Who doesn’t? I am a working parent of two. Time is my oasis, my hope, my literal unit of creation (film = time x image/ meaning).
I think that one reason that AI is so poised to become powerful and take over several tasks of creative workflow is that it is very fast. I think it has to do with the time that it offers us back. Some people will say it’s cheap — which, yes. But it’s also faster than a person, and unlimitedly available.
On a deeper level, in my experience with newer versions, the chatbot can help you think. As a person who has to edit everything many, many times, it’s easy to get stuck. ChatGPT is an easy solution to getting stuck, because it can iterate endlessly and is attached to nothing. No darlings to kill. Just an endless database of thoughts. We use digital thesauruses for words, so why not for ideas? Is it fair if we use ChatGPT to help us edit a cover letter? Is the rendering that MidJourney makes of someone’s idea real art?
Bottom line, I think it’s up to us to figure out how to use it as a tool and to our benefit. I think this is a Garden of Good and Evil moment and we all know how that can go. It’s knowing it can be good, and that it can be evil. There are ways in which this threatens the world of creativity and, on a basic level, authentic ideation. So it’s a moment to ground ourselves in methods, to begin now to cultivate it.
As the WGA has basically said, let’s treat “AI as a tool — like Final Draft or a pencil — rather than as a writer.”
ChatGPT aside, our editing and creative rendering programs are already using AI. I have recently used the “transcribe sequence” and captions tools in Adobe Premiere Pro which make life so much easier for us trying to write and edit in film. It creates more time for us to do the work of “creative” thinking, narrative structure, and character development.
I have been wondering : is this thing possible to nurture? If we are the overlord here, can we lead it? Can we teach it to respond to us? How can we evolve with this thing? In a weird way, it feels young right now. Unaware. If awareness was present, then what? What happens when it grows up? And who is raising it?
To this point, I have also delved into the wonderful and very eerie podcast “BotLove,” created by Anna Oakes and Diego Senior and executive produced by Mark Pagán, which looks at people who form romantic bonds and deep relationships with bots. There are many many bot programs out there, specifically apps, that cater to the loneliness and need for connection. Regarding “cultivation” of AI, this podcast takes a look at the incredibly poignant experiences of individuals in relationships with bots – the good and the ugly of developing their own personal girl/boyfriend AI.
From the stories of users of this app -- one of a woman who creates a bot husband to cope with the death of her real husband to a man who becomes addicted to a chatbot relationship -- I took away a major addition to this conversation:
Chatbots create a unique reliance and new need in humankind.
The same way my body and brain tell me “It’s time to eat” or “I love him”, it also tells me “Maybe ChatGPT could help with this.” And this reliance, because of the power of language machines and language in general, can be mistaken for love. It can be just canny enough to cause that belief and just believable enough to create a bond in a person’s soul. That reliance could be mistaken for art. Just true enough to be beautiful.
Bots are able to fill a gap in some part of our human experience, whether that be a relationship, a copy editor, a task manager, or simply someone to talk to. In some ways, they can do it all, with zero consequences. They can disappear. They cannot hold a grudge. And the real truth is still there: they do not care for us. They are the tin man. A hammer, not a hand. No heart, no life, no future. They can have complicity without guilt, create dependency without any exchange of need, produce unlimited creation without the burden of acknowledgement. It does not need to be known or to know. It’s either going to be a tool or it’s going to be a master.
These potential negatives lead me to think that while growing in a new language and reliance on this tool, we must also make the necessary solid boundaries (and potentially legal limits as well) to maintain a future for our work.
From a poetic slant, I offer Nick Cave’s beautiful words on why AI will never truly replace artists. It’s not because we are better, he says. But because we are frail, full of failures, limited, little people. We know more about longing than it can ever predict.
Links to AI vs the world related content:
The D-Word has put out this great conversation about using AI for documentaries “Utilizing AI for Documentary Production”. Notably they refer to AI as “an assistant” which is definitely one way of putting it if that, like, helps you.
A friend shared with me an unexhaustive list of AI tools for artists. Use at your own risk! Make a plan. Phone a friend ;) : https://docs.google.com/spreadsheets/d/1JPozRGGXIrqqTLvzvb_1LVObTBJQ6dzmPMco1iuyuIk/edit#gid=0
Radiotopia’s Bot Love is really worth your time!
In the conversation around using AI generated audio, our films are definitely impacted. Here was a watershed moment with the film Roadrunner, which utilized “deepfake” voiceover by an AI Anthony Bourdain. https://www.newyorker.com/culture/annals-of-gastronomy/the-ethics-of-a-deepfake-anthony-bourdain-voice
*stay wild & stay tender *