“I choose to spell out a positive vision about the possible impact of AI on humans because there is already a lot of negative commentary – much of which I agree with. Still, I think we can hope that the changed way AI helps humans see the world will be in valuing the particulars and the truths that AI and machine learning unearth. That will stand in contrast to humans’ longstanding efforts to try to create general truths, laws and principles.
“General ‘laws’ humans have theorized about the universe teach us a lot. But they can be imprecise and inaccurate because they don’t account for the wild mass of particulars that also point to truth. We humans don’t have the capacity to ‘see’ all the particulars, but AI does.
AI/machine learning tools are better equipped than humans to discover previously hidden aspects of the way the world works. … They ‘see’ things that we cannot. … That is a powerful new way to discover truth. The question is whether these new AI tools of discovery will galvanize humans or demoralize them. Some of the things I think will be in play because of the rise of AI: our understanding of free will, creativity, knowledge, fairness and larger issues of morality, the nature of causality, and, ultimately, reality itself.
“Here’s an example: In 2022, researchers discovered we have the ability to predict heart attacks amazingly accurately after they ran a small data set of retinal scans through an AI analysis system. It turns out the power of simple retinal tests to predict heart attacks was unexpected and often better than other tests had demonstrated.
“We don’t know exactly why that is, but the correlations are strong. A machine system designed to look for patterns figured it out without being told to hunt for a specific thing about the causes of heart attacks. This use of artificial intelligence turns out to be much more capable than humans at discovering previously hidden aspects of the way the world works. In short, there is truth in the particulars and AI/machine learning tools are better equipped than we humans are to discover that reality. AI tools let the particulars speak. They ‘see’ things that we cannot and do so in a way that generalizations don’t. That is a huge insight and a powerful new way to discover truth.
“Now, the question is whether these new AI tools of discovery will galvanize humans or demoralize them. The answer is probably both. But I’m going to focus on the positive possibilities. I’m convinced this new method of learning from particulars offers us a chance to rethink some of the fundamental ways we understand ourselves. Here are some of the things I think will be in play because of the rise of AI: our understanding of free will, creativity, knowledge, fairness and larger issues of morality, the nature of causality, and, ultimately, reality itself.
“Why can we reimagine all those aspects of life? Because our prior understanding of them is tied to the limits of our brains. Humans can only think about things in a small number of dimensions before problems get too complex. On the other hand, AI can effectively function in countless multidimensional ways with an insane number of variables. That means they can retain particulars in ways we can’t in order to gain insights.
One idea that could come back in this age of AI is the notion of causal pluralism. Machine learning can do a better job predicting some causal incidents because it doesn’t think it’s looking for causes. It’s looking for correlations and relationships. This can help us think of things more often in complex, multidimensional ways. … I am opting for a very optimistic view that machine learning can reveal things that we have not seen during the millennia we have been looking upwards for eternal universals. I hope they will inspire us to look down for particulars that can be equally, maybe even more, enlightening.
“Let’s look at how that might change the way we think about causality. Philosophers have argued for millennia about this. But most people have a common idea of causality. It’s easy to explain cause and effect when a cue ball hits an eight ball.
“For lots of things, though, there really can be multiple, reasonable explanations of the ‘cause’ for something to happen. One idea that could come back in this age of AI is the notion of causal pluralism. Machine learning can do a better job predicting some causal incidents because it doesn’t think it’s looking for causes. It’s looking for correlations and relationships. This can help us think of things more often in complex, multidimensional ways.
“Another example can be seen in the ways AI and machine learning might help humans advance creativity and teach us about it. Many creative people will tell you that when they are creating they are in a flow state. They did not start the creative process with a perfectly clear idea of where they’re going. They take an action – play a note, write a word or phrase, apply a paint brush or … my favorite example … chip away at the rock because the figure to be sculped is already in the stone and just ‘waiting to be released.’ Every time they take that next step they open up a new field of possibility for the next word or the next brush stroke. Each step changes the state of the thing.
“That’s pretty much exactly how AI systems operate and try to improve themselves. AI systems are able to do this kind of ‘creative work’ because they have a multi-dimensional map – a model of how words go together statistically. The AI doesn’t know sadness or beauty or joy. But if you ask it to write lyrics, it will probably do a pretty good job. It reflects our culture and also expands the field of possibility for us.
“Ultimately, I am especially interested in ways in which this new technology lights up the world and gives us insights that are enriching and true. Of course, there’s no great reason to think that will happen. Computers have lit the world in ways that are both beautifully true and also demeaning. But I am opting for a very optimistic view that machine learning can reveal things that we have not seen during the millennia we have been looking upwards for eternal universals.
“I hope they will inspire us to look down for particulars that can be equally, maybe even more, enlightening.”
This essay was written in January 2025 in reply to the question: Over the next decade, what is likely to be the impact of AI advances on the experience of being human? How might the expanding interactions between humans and AI affect what many people view today as ‘core human traits and behaviors’? This and nearly 200 additional essay responses are included in the 2025 report “Being Human in 2035.”