Medicine May Be an Art, But AI May Be Artists

Six hundred years ago, Swiss physician/scientist/philosopher Paracelsus disclaimed: “Medicine is not only a science; it is also an art.”  Medicine, most people in healthcare still believe, takes not just intelligence and fact-based decision-making, but also intuition, creativity, and empathy.  This duality is often cited as a reason artificial intelligence (A.I.) will never replace human physicians.

AI-Da and her self portrait.  Credit: Mark Thomas/Rex/Shutterstock

Now, I have to admit, “she” wasn’t on my radar either until recently, when she was imprisoned/impounded at customs by Egyptian authorities on her way to an art exhibit at the Great Pyramids of Giza, where she was scheduled to show her work.  Egyptian authorities first objected to her modem, then to the cameras in her eyes.  “I can ditch the modems, but I can’t really gouge her eyes out,” said her creator Aidan Meller.  After a 10 day stand-off, she was released late last week. 

Perhaps those skeptics have not heard about Ai-Da. 

Let me back up.  Named in honor of famed 19^th century mathematician/programmer Ada Lovelace, Ai-Da is “the world’s first ultra-realistic humanoid robot artist.”  She was created in 2019, and uses AI algorithms to create art with her cameras/eyes and her bionic arms.  She can draw, paint, even sculpt, and had her first major exhibit – Ai-Da: Portrait of the Robot -- this summer at London’s Design Museum.

The description of her exhibit says:

As humans increasingly merge with technology, the self-titled robotic artist, Ai-Da, leads us to ask whether artworks produced by machines can indeed be called ‘art’…Ai-Da can both draw and engage in lively discussion…These features, and the movements and gestures that Ai-Da is programmed to perform, raise questions about human identity in a digital age.

Her website elaborates:

…current thinking suggests we are edging away from humanism, into a time where machines and algorithms influence our behaviour to a point where our ‘agency’ isn’t just our own. It is starting to get outsourced to the decisions and suggestions of algorithms, and complete human autonomy starts to look less robust. Ai-Da creates art, because art no longer has to be restrained by the requirement of human agency alone.  

Here's a video: 

Lest anything think Ai-Da is a one-off, I’d also point to Xiaoice, a Microsoft-built, China-based AI chatbot that is “a poet, a painter, a TV presenter, a news pundit, and a lot more.”  Microsoft spun it off in 2020, the company maintaining the name while renaming chatbot Xia Yubing.  Xia is now creating traditional Chinese paintings, having already mastered Western-style painting during its Microsoft time. 

Xia appears to have passed an art version of the Turing test; according to China Daily: “In 2019, works of art produced by Xia were submitted for an exhibition of postgraduates' work at China's Central Academy of Fine Arts. When Xia's paintings were presented beside those of humans, nobody realized they were generated by AI.”

There are other AI artists besides Ai-Da and Xia.  London had “the first international AI art fair” this month – deeep, featuring “the world’s largest collection of AI created art.”   One reviewer found the works “are equal parts hypnotic, unsettling, and produce an outlook quite alien to traditional styles.”  If step one for AI art is to create art that we can't distinguish from human art, then step two is to create art that only AI could create.   We may already be there.  

If AI-produced art isn’t impressive enough, earlier this year AI was used to finish Beethoven’s famous unfinished 10^th symphony, synthesizing all his other works and his notes, and using them to create something he might have written.  It succeeded: “We challenged the audience to determine where Beethoven’s phrases ended and where the AI extrapolation began. They couldn’t.” The completed symphony had its world premiere earlier this month. 

If you'd like to listen:

In fact, for all you know, this article could have been written by an AI, such as Rytr, which, according to The Next Web, “brings the skill of a talented freelance writer to the digital realm, generating copy that can give flesh-and-blood writers a run for their money.” 

Healthcare certainly hasn’t been ignoring AI.  Every day it seems there are more announcements about AI-powered innovations, as well as funding for AI-based companies with a health focus. 

Just week, researchers at the University of Utah Health/Rady Children’s Hospital reported they’d used AI to parse massive amounts of genetic data to diagnose rare pediatric disorders, in a way humans never could have.  AI is already also increasingly important in drug discovery, and numerous health systems are implementing their own AI-based initiatives, such as a Stanford University Medical Center/Microsoft project on medical imaging datasets and a Mayo Clinic/Google AI algorithm for treatment of brain diseases. 

Last year alone the FDA approved 100 AI/ML (machine learning) devices, with radiology being the big leader, according to a Politico analysis; as Dr. Eric Topel likes to say, it is the “sweet spot of AI.”   

Lenovo’s Sinisa Nikolic believes: “AI is set to transform the future of healthcare,” although he offers the usual cautions: “In all aspects of healthcare, you will always need human-human contact and interaction. Humans have empathy; machines cannot replace that. AI will help us be better, stronger, and healthier.” 

Not everyone is as conservative.  Kai-Fu Li, author of AI Superpowers, predicts: “I anticipate diagnostic AI will surpass all but the best doctors in the next 20 years.” 

Healthcare has come a long way with its acceptance of healthcare, from initially rejecting it, of course, to the now common mindset that, yes, it could be a great help, helping automate common tasks and augmenting clinicians. But crossing that line between augmenting and replacing is hard for many to accept.     

We can accept AI being good at the “science” part of medicine, but we’ve yet to be convinced it could be good at the “art” part of it.  But, as Ai-Da and other AI artists are illustrating, it’s something we’re going to have to face. 

Ai-Da’s website warns: “If Ai-Da does just one important thing, it would be to get us considering the blurring of human/machine relations, and encouraging us to think more carefully and slowly about the choices we make for our future.”  

Let’s hope healthcare thinks carefully – but not too slowly -- about the choices AI offers us for the future.