A man sits at a desk with his feet up, surrounded by computer programming and philosophy books. He looks out the window at the bare branches of a tree behind a sheet of winter blue sky. In his reverie, he is haunted by a question that he can’t seem to answer: In the age of Artificial Intelligence, what is intelligence? He has spent his life working with artificial intelligence, but as AI has seemingly, overnight, skyrocketed into the forefront of cultural consciousness, he muses more than ever on his life’s work. Outside, the blue turns to deep orange, and the lights begin to shimmer in the city. He packs up his things and leaves his office, relishing the cold air against his face. As he walks down the university steps, passing the skaters, the brightly-haired students, headphone-wearing joggers, the faded concert and art posters on telephone poles, the answer to his question hangs just out of reach. While surveying the humanity that surrounds him, he knows there is something, something about intelligence, that can’t be manufactured.
Michael Littman is a computer science professor at Brown University and serves as the division director for the Information and Intelligence Systems at the National Science Foundation. Timothy Henry is associate professor in the department of computer science and information systems at Rhode Island College. Littman has been working in the AI sphere for 40 years, referring to himself as a “cognitive science groupie.” Henry is currently on the board for developing a new AI major at RIC and was a project manager who helped apply cutting-edge technology to law enforcement, first response, healthcare, and education. They are both experts in their fields, and as they watch AI blossom into mainstream vernacular, they find themselves preoccupied with the same question: What is AI’s place in humanity’s future?
Littman and Henry’s technical definition of AI is something that isn’t distinctly definable yet, like our intrinsic grappling with the concept of existence, or consciousness. Henry says, “That’s one of the challenges, there’s no real good definition because there’s no real good definition of intelligence,” but its main job is to reason like a human. Littman specifies that AI “at the end of the day is about creating software that solves problems.”
We interact with AI every day, and we may not even know it. For those of us who have an iPhone, there is a recent feature that uses AI to follow your daily patterns, and offer directions based on your usual daily destinations. Henry elaborates that AI “is based on probabilities. It has learned from the past what’s going to happen in the future.” AI is an algorithm of putting data in to get data out, or as Littman puts it, a connection between inputs and outputs.
To make sure AI doesn’t begin to model the wrong side of history, it is incredibly important that the people who control the inputs are on the right side of history. Henry defines it as “garbage in, garbage out.”
Although some AI data contributors are working to eliminate biases from AI by making sure to train it in equality, there have been instances of bias in the technology. For example, there have been court cases in which AI is used to assist a judge in sentencing people committed for crimes when it projected longer sentences for people of color or lower incomes. If AI’s existence was created as a reflection of human intelligence, it has the capacity to emulate the best and worst parts of us; just like a young mind has the capacity to learn how to think critically, or to think narrowly, depending on what it is taught. Which is why, in the beginning of AI’s introduction into our society, the way it is taught will define how it acts.
Littman passionately opposes the assertion that artificial intelligence is intelligent. “I would say that we have not achieved artificial intelligence. I wouldn’t put it in the past tense, we haven’t done it yet.” Littman uses the example of chatbots like ChatGPT as an example for the inability of AI to be intelligent. He says that these chatbots are just “words on words. They can reflect on things other people have written. The machine isn’t having these experiences itself.”
Our biggest mistake would be to put it on par with the intelligence of a human being. Henry cautions to remember that it is a tool that is great at helping make decisions. “There are two things it does, it can make decisions and perform actions. The decision it makes is a recommendation and you take action. Or it makes a decision, and you give it permission to take the action itself. We have to build trust in our AI before we allow it to make the decision and take the action itself.”
On the edge of a technological boom that spawned recent advancements in AI, the future seems daunting, and vaguely uncertain. We all question the capabilities of AI, fearfully amusing ourselves with the dystopian, or a vibrant society of Jetson-esque reflection. Henry approaches the question of the future with an air of pragmatism; he relates our acceptance of AI to the original reluctance to share information about ourselves on social media. “People were appalled at how much social media and Google has captured data about us, but a lot of people don’t care because they like the services. Something similar will happen with AI. The more information you share about yourself, the more able it will be to help with your tasks.”
The future, to Littman, still holds humanity’s essential questions as key. He doesn’t subscribe to the transhumanism idea, that machines are going to replace us. “People are still going to be people, the things that drive us, our loves, our frustrations, and our bitter rivalries, that’s going to stick with us.” And in the far, far future, long after us? “It’ll still be people doing stuff, we’ll just have cooler things. Maybe that’s a boring future, I don’t know. But I think that’s really reassuring, we are who we are, and we’re on this ride together to try to figure out how the universe works.”
