Artificial Intelligence Can Learn From Seniors
The benefits of broad input.
Posted March 31, 2022 | Reviewed by Vanessa Lancaster
- Artificial intelligence is built on the assumption that machines can learn. In learning tasks, machines can even teach each other.
- Elders are the repository of knowledge and can add much value to AI creations.,
- AI creation needs to include various learning approaches and ethical and philosophical input, a hallmark of elder thought.
All too often, when new technological advances make their way into the mainstream, they benefit young adults, who are generally their creators. Sometimes children are the beneficiaries but seldom are elders considered, which matters to me, an elder psychologist. There’s actually an evolutionary reason to explain that.
The ultimate goal of any species is to replicate itself and keep the species alive. So from an evolutionary perspective, adults of reproductive age and their offspring are valued most highly. There has never been an evolutionary role for the long-lived members of a species because they are beyond the reproductive stage of life. But most humans and some animals live well beyond childbearing.
Today, many adults in developed countries can expect at least seventy years and maybe even one hundred.
The Benefits of Senior Input
What is our role then? Certainly, we must provide some benefit to human evolution. And we do. As elders, we collect and transmit our species’ cultural rules and history. Since the beginning of recorded time, we've done this—consider shaman or priests’ stories passed on by word of mouth before written language.
Cave drawings communicated hunting rituals and important ideas worth remembering about dangers from nature or other tribes. We, elders, are the repository of knowledge, with brains full of information and how-to skills accumulated over our decades of living. As a generation, we are the experts in problem-solving, able to select new connections between ideas and new ways of doing things from our vast knowledge base.
Wisdom Adds Value to AI Perspective
While older folks don’t have the physical agility or energy of young and middle-aged adults, we make up for it with wisdom—the ability to draw on our accumulated know-how and create novel solutions to everyday problems. On the other end of the age spectrum, young children definitely have creative abilities and haven’t yet internalized learning biases that limit the way they will eventually think. But elders have a unique perspective on how the world works.
The current environment in which we live is increasingly dependent on technologies that depend on artificial intelligence (AI). And in my attempt to understand more fully where elders fit in the realm of AI learning, I attended the “Technology, Mind and Society” conference sponsored by the American Psychological Association. Not surprisingly, there was very little focus on elders.
What Is AI and How Does It Work?
Artificial intelligence is built on the assumption that machines can learn. In learning tasks, machines can even teach each other and further the subsequent knowledge base of new machines. By creating algorithms based on huge amounts of data, machines can analyze and solve problems much more rapidly than human intelligence ever could.
Applications of AI permeate just about every field, from manufacturing to gaming and almost every facet of life. Your smart refrigerator remembers that you are almost out of milk and ambient temperature controls on your thermostat learn the settings that keep you comfortable.
AI historically has been in the bailiwick of engineers, young men. They create algorithms, write code to manage and direct the algorithms, and develop products and services based on them. But if we rely entirely on young males, AI will be skewed by their biases or limitations in thinking characteristic of their stage in life.
Toward a More Human AI
At the "Technology, Mind and Society" conference, Alison Gopnik, a professor at the University of California, Berkeley, discussed what AI could learn from four-year-olds. She discussed young children’s cognitive flexibility, meaning their curiosity-driven ability to see new patterns or make novel generalizations because young children's brains are very plastic, open to all possibilities. Four-year-olds aren’t worried about making mistakes and don’t avoid risks.
The learning-trap sets in with adults who avoid certain risks or fear making mistakes, according to Gopnik. So while the young child’s learning approach is to explore, the adult’s way “exploits” their knowledge. On the other hand, Elders neither explore nor exploit in their characteristic approach but provide an overview and a context for knowledge—passing on big chunks of culture as did elders from the beginning of human time.
There’s room for all three learning approaches in creating artificial intelligence, but currently, the very young and the old have little input into AI thinking. Broadening AI creation to include various learning styles, as well as the ethical and philosophical input that is a hallmark of elder thought, might keep our society more human, less robotic.