I remember a slumber party as a kid when my friend's mother dressed as a fortuneteller, all scarves and bangles in a mystical basement setting, and individually told us details of our own lives with spine-tingling accuracy.
How could she know these things? Was it intuition? Or was she just good with a telephone? It turned out she had called our parents ahead of time and gathered fun, real-life details to sprinkle into our fortunes, with enough accuracy to spark our amazement and wonder.
Which brings me, in a roundabout way, to artificial intelligence — how it also predicts, weighs probabilities, has access to information we can't always see, and sounds certain even when it's unsure.
To show the evolution of computing, the Computer History Museum in Mountain View, California, guides visitors from the abacus — the ancient, hand-operated calculating tool — to Ameca, a robot with human-looking features created by Engineered Arts that uses OpenAI's GPT-4 to chat with patrons and answer their questions. Ameca can shift between languages and mimic human expression. Ameca is a marvel, and when I asked it to identify the scariest robot, it suggested the Terminator. "Oh no, Ameca, I think it might be you," I told the robot. Ameca tried to reassure me.
Ameca uses AI, "the ability of a computer system to perform tasks that normally require human intelligence — such as learning, reasoning, perception, decision-making and language understanding." (A definition I just got from querying Copilot, Microsoft's AI chatbot.) Ameca is just perhaps a more whiz-bang version of the technologies that are becoming increasingly common in homes and workplaces with access to technology.
Viewed one way, AI allows us to experience a little of the magic from that long-ago slumber party. The models generally have scraped data from a variety of sources — prompting a debate of its own about content creation, ownership and appropriate use — and can now mine and analyze data in ways few of us could previously have predicted. It doesn't just follow rules; the technology learns, adapts and sometimes acts.
For Catholic healthcare, and at CHA, these developments have led to a variety of ethical, spiritual, moral, medical and legal questions, debates and learnings. Several, but not all, are explored in this issue of Health Progress centered on AI. It's not just a discussion — we've been doing that. This issue starts to shape responses to significant questions about the technology: how it should be considered, used and not used.
There's more to explore in the future, including how to ensure more automated billing practices respect the dignity of patients, how to weed out human bias from large datasets used for population health findings, and how to protect mental health and promote healing in a world where technological innovation should support human tasks, but not overtake human connection.
It's hard to believe that it was 1954 when mathematician Norbert Weiner wrote, "If we, in a small way, make human tasks easier by replacing them with a machine execution of the task, and in a large way eliminate the human element in these tasks, we may find we have essentially burned incense before the machine god. There is a very real danger in this country in bowing down before the brass calf, the idol, which is the gadget."1
The promise of AI is staggering. So, too, are its implications for misuse, misplaced allegiance or idolatry. Much like early mistrust of the use of the abacus, or my own reservations about Ameca, it will be the human hand in testing, guiding and constraining AI that will chart its path. And one key question in the decision-making will surely be: Is this for the greater good?
NOTE
- David A. Hollinger and Charles Capper, eds., The American Intellectual Tradition, Volume II, 1865 to the Present (Oxford University Press, 2016), 379.
You'll notice that several authors reference Pope Francis' Antiqua et Nova, and not Pope Leo XIV's Magnifica Humanitas. Articles for this issue were written and edited prior to the release of Pope Leo's encyclical on AI.