Two sides of AI

By: Courtney Garvin / January 27th 2023 / Design
Two sides of AI

Artificial Intelligence (or AI for short) has been all the buzz the last half of the year. Experts with knowledge of the incredible technological leaps voiced that the public had no idea what was in store.

Then in 2022, the public gained access to these powerful tools—Dall-E (and subsequently Dall-E 2), Midjourney, Stable Diffusion, and ChaptGPT3, to name a few. These futuristic tools allow users to enter text prompts to generate visual and written content. Social networks have been abuzz with people showcasing their AI co-creations. They have even expanded to find additional creative ways to use these tools—from creating children’s books to writing music or, my favorite, my brother’s genius idea to have ChaptGPT3 play as dungeon master in a Dungeons & Dragons game with his kids and him while Midjourney generated the corresponding visuals.

But truth be told, the power of AI has already been in use before these new products became available. For example, customer service bots and grammar-checking services such as Grammarly have already become part of our daily workflow. In addition, users of Adobe software products have been benefitting from Adobe’s powerful AI engine, Sensei, for some time now to manipulate images or sound. But these new AI tools take these concepts to the next mind-blowing level.

With any new technology, the long-term consequences and ethical ramifications often lag behind. For example, as soon as these new tools were unleashed, people raised red flags on the ethics of how the models were trained and whether or not content was created using data without consent of the creators. Later, it was revealed that, in fact, some models were trained on artwork without the artist’s permission. These issues are still being hashed out in the public forum, and any potential legal ramifications are still yet to be decided. So how should we view these new tools and their awesome — in that they invoke both awe and fear — power? One way to look at the value of AI is through the lens of creative tools.

Creativity is the ability to make something new of value. We need to see beyond the expected or routine to have a creative idea. But that is the dilemma. How can we break outside the constraints of what we already know or have seen before? Especially, how can AI trained in rote unthinking mathematical algorithms generate insights into the human condition? One way to solve this problem is to look at how others have solved similar problems in the past.

Musician Brian Eno and artist Peter Schmidt sought to find a remedy to the creativity problem in the 1970s. Together, they created Oblique Strategies, a deck of cards, each with a prompt—some seemingly on topic and others out of left field. Artists, looking to break their creative block, shuffle the deck and draw a card. They use the random prompt as a jumping-off point to think about their problem differently. The true value of Oblique Strategies is not the actual prompts—some, like “A line has two sides” at first glance, may seem unrelated to the task—but rather the ability to interject newness into our thoughts or break hidden assumptions. Like Oblique Strategies, AI tools can help us see or try new things with minimal up-front time or materials investment. We can explore more, try more, and maybe make unusual connections that are the seeds of creativity.

The consequences of these new AI tools are yet to be determined. Some have used these tools to reach new heights of creativity by quickly generating fascinating visuals along with deeply moving texts whilst others have found that the AI is often mired in the uncanny valley generating image after image of creepy disturbing visages that keep reappearing (for example the case of “Loab”, the woman that keeps appearing when using negatively weighted prompts into text to image AI).

Still, it is worth exploring all of the potential uses and benefits. History can also help us put these advances in context. The advent of every innovative technology has had its enthusiasts and opponents. Many have sighted the advent of photography as a similar disruptor to the new AI tools. With photography, people predicted the end of painting or art. For why would anyone paint a picture when they could reproduce the actual image almost perfectly? But photography was not the death of painting and, instead, became its own unique artform. AI-generated art may follow the same path: not displacing previous art forms but creating a new way to create based on the fusion of computers and humanity.