New Video: How will Multimodal AI models like DALL-E Impact Society?
If you haven’t noticed already, the series is rapidly shifting towards a broader discussion about the implications of Multimodal AI. In addition to the creative lessons shared up until this point, it’s important to put the series in a broader societal context and discuss these implications as a group. I hope you find the rest of the series intellectually stimulating and also challenging. I’m interested in hearing everyone’s opinions and expanding our vision of what could happen.
Today’s video speculates on the impacts multimodal AI will have on our society. What can we imagine could happen? How will our world change if everyone can access these tools?
YouTube Transcript (SPOILER WARNING)
Let’s speculate for a bit here. How will AI generating different kinds of art impact our society?
The biggest consequence I can foresee is simply causing us to debate the value of human vs. AI creation in the first place. Is AI generated art considered cheating? On the one hand, purists may argue that art coming from a machine does not come from the same limitations as a human, making it worthless. Humans have a limited time on this earth and unique life experiences biological, environmental, intergenerational, or otherwise making our creations more valuable, whereas a machine can theoretically generate infinite works of art without having lived any of the human experiences that actually drive our emotions. Some purists would even go as far as saying that human art is ineffable, at a level that AI could never understand itself let alone generate with the same charm and divine character.
At the same time, since multimodal AI can bridge the skill gap required especially compared to old school tools like Photoshop, my suspicion is more people will be creating art more often. Imagine using DALL-E regularly to send endearing, personalized greeting cards to all of your friends and family for every occasion.
This new found ability to easily generate art in the first place, may actually expose people and get them interested in creativity and the digital arts. So, one thing that GPT-3 actually suggested to me while writing this script, was that maybe multimodal AI will actually increase society’s appreciation of art. Maybe creativity will finally become something in the future that everyone could participate in and begin to take seriously. This sort of meta-appreciation for art and creativity, human, machine generated or otherwise could spark a new school of thought around the importance of art and creativity in the first place. Which is a really, really interesting societal effect to think about.
Finally, I’m excited for multimodal AI to question our core human beliefs around the classic educational subjects. We tend to believe science, technology and art exist on different ends of the spectrum. One of my favourite documentaries, Tim’s Vermeer, explores the idea of reverse engineering tools that the famous painter Vermeer may have used himself personally to create his great works of art. In my opinion, Tim created some pretty impressive replicas just by focusing on the science side of Vermeer's work. One lesson I took away from the documentary is that during the renaissance era, art and technology were viewed as the same thing. Using technology and tools was just a part of your everyday art and creative practice. I’m excited for the possibility of moving to a new world, one where society appreciates them both. There's no reason why these kinds of AI models couldn't generate math theorems, potential scientific research studies, or novel patent ideas. Perhaps, multimodal models can provide a bridge between the different educational subjects and the people behind them, with a lot more cross pollination. This idea allows for a more interdisciplinary, humanitarian, inclusive, but also a radically progressive society.
The Key Idea
I’m really scratching the surface here, but I would say that altogether:
multimodal art will likely cause a debate on the nature of creativity itself
by greatly reducing the barriers to create new things, the creative class of people may grow larger in size, leading to a greater societal appreciation for art overall
Finally, multimodal AI may connect various human disciplines together more, further blurring the lines between art and science