By Jack Cumming
The woman who is sometimes called the godmother of artificial intelligence, Fei-Fei Li, began her AI journey with an insight. She observed that before the evolutionary explosion, what life existed on Earth was blind. 538.8 million years ago there was a sudden burgeoning of creation known as the Cambrian explosion. The trigger? That was the evolution of eyesight and vision.
The Wonder of Eyesight
Of course, those people who are born blind depend on the rest of us to describe the visual world for them. Imagine what life would be like without sight. For sighted people, most of our experience of the world, especially our earliest experiences as infants, comes from vision. Body language often reveals what words conceal.
That Cambrian insight gave Fei-Fei Li a brainstorm. If we want computers to mimic human understanding, we have to give them vision, images, and analytical algorithms to recognize those images. Her thought wasn’t popular at the time. She was warned she might not get tenure if she persisted, but she did, to our benefit.
Brain and Computer Science
Fei-Fei’s graduate program at Caltech combined neuroscience with computer science. It was natural, then, that she saw the value of neural networks (nodes and connectors like the structures in our brains) as a way to train computers to recognize images. The result was a massive project called ImageNet.
This breakthrough has immense implications for all our lives, including the lives of workers and residents in senior living. It’s not up to our vendors to come up with AI wonders to sell to us. It’s up to us to specify what we want computers to be able to do for us. As Daniel Huttenlocher of MIT reminds us, “AI does not have human values, intention, motivation, emotion, morality, or judgment” … and we might add that it also lacks grace, mercy, or discretion.
Huttenlocher describes AI as an amplifier of human behaviors. Humans can do research or use imagination to improve the human condition, or humans can unleash violence for self-serving purposes, as in war. AI can “amplify” both intentions, but the humans behind AI are the ones responsible. Click here for more on the governance of AI.
AI Is Not Human
Moreover, AI only mimics human emotions; it doesn’t feel them. Is that emotional mimicry misleading and manipulative? Perhaps so, or perhaps it makes interacting with computers more acceptable and satisfying for users. We all know the frustration that comes with an endless telephone tree (press one for sales or press two for billing).
AI has the potential to shortcut your way through that tree while giving you the feeling that the machine-person is actually trying to help you. It would be best, though, to at least know that you’re talking with AI and not just a person in a call center overseas. That seems like a minimal ethical disclosure.
But It Sure Can Help
Just imagine how having a gifted conversational assistant could help you in fielding repetitive questions from prospects or from residents. Think of a prospect calling to ask if you can immediately accept the prospect’s mother who is recovering from a stroke and whose physician has advised to no longer live alone. AI can answer the caller and free you to be productive.
Or, imagine augmented intelligence fielding a call from a resident who wants to submit a lunch order (“What’s the menu today?”) or a call like this: “I took out the trash, and I forgot my key. Can you, please, unlock my door?” The potential of AI (augmented intelligence) is particularly vast for a people-dependent, people-serving business like senior living.
Take the Initiative With Vendors
If we just ask our vendors, “How are you using AI, and will you develop applications with us?” providers may miss the opportunity, much as the value opportunity for resident-serving intranets was missed for over a decade. The onus is on industry technologists to educate themselves to the potential and to take the initiative to ask for what can help.
Vendors may know computers, but you know senior living. Learn enough of what computers might do to be able to ask intelligent questions and to make intelligent requests. There is a wealth of educative material on YouTube and elsewhere on the internet. Master the concepts to avoid being hoodwinked.
Example: Imagine error-free delivery of medications. What would that be worth in liability avoidance, not to mention the improvement in trust as your brand becomes known for excellence? The applications of AI are manifold, especially if the obvious marriage of AI with devices (including robots) is consummated.
Just Think of a Resident’s Television
We’ve noted elsewhere the potential that a television equipped with an adequate camera can be for combatting resident loneliness and enhancing resident safety and communications. Think vision. A camera-equipped TV can reassure residents that help will be there when they need it, or it can observe residents to help them with reminders and more.
Imagine what that could be if the television were equipped to respond intelligently and immediately with what is observed. Not only can AI improve resident monitoring, but it can do so while retaining resident dignity. Being chided by a machine to take your medicine is much less humiliating than having a person do the same.
Additionally, a camera-equipped television can bring a resident into the homes of distant friends, children, and long-lost relatives. That can be a godsend for a lonely resident who no longer fits in well into the lighthearted chatter of the dining room, or who no longer has the finger dexterity for the crafts like sewing and woodworking, mainstays of life enrichment. Many residents just withdraw into the solitude of their rooms as the bustle of the public areas morphs into confusion for an aging mind.
One wonders why industry leaders didn’t insist on more after Amazon worked with TCL to introduce a television that responded like an Echo Show to voice commands. Many televisions require that voice interactions be uttered through a push-to-talk button on a TV remote. Not the Amazon television. The Amazon TV is much more senior savvy. But it lacks a camera. Virtually every laptop has a camera. Televisions don’t.
Why didn’t the senior living industry step forward to encourage Amazon to at least include a built-in camera and HIPAA capability for tele-visits with physicians, family, and other interactive operations? Such an AI-enabled television could be particularly welcome for those poor souls who are bed-bound for the rest of their lives in skilled nursing facilities. We can only hope that this is not a totally lost opportunity.
Visionary Seeing
As Fei-Fei Li perceived with such profound insight, seeing is the key to the intelligence which brings us so many advances. Cameras are the eyes of the machine world. People bring the responsibility essential to the moral and ethical guidance of machine deployments. AI combined with creative engineering imagination promises to bring us products to make old age in the future what we wish we could provide today.
Those who are now in their 40s and 50s will be the next generation to age into dependency. Let’s hope that senior living takes the initiative to ensure that these evident advances are widely available to give their lives meaning and to offset debilities of old age, thereby empowering those now working in senior living to have fulfilling lives later when they are residents.
