“Two years ago, my prediction was that humans would use AI with a mixture of rapture and horror. While ‘horror’ may be an overstatement, ‘concern’ may be increasingly appropriate: A 2021 Pew survey showed that 37% of US adults were more concerned than excited about AI; by 2023 that number had grown to 52%. My prediction (and an easy one at that) is that number will continue to grow. I find that even many of my Carnegie Mellon colleagues are what I would call ‘suspiciously optimistic,’ overall positive, but let’s just keep an eye on this.
“Right now, my colleagues and I are embarking on a research project that couldn’t be done without AI. It will see if AI can be a change agent that, using evidence, can talk you out of a false belief. That sounds promising, but what happens when people realize that it wasn’t just science, wasn’t a human correcting a wayward view, but was AI? Will they feel played? Misled? Victimized? Will they be angry? Or thankful? Will AI be seen as a human surrogate – a friend gently guiding us to truth – or something more sinister? Does it take away agency or add to it?
If AI in fact eventually achieves consciousness, then what? Suddenly it changes the nature of how we define what it means to be human. Who will feel more existential dread then? Us – of the AI – or the AI – of us? How then does that impact feelings of happiness or sadness, meaningfulness or ennui, psychological richness or abject pointlessness?
“When we live in a world with AI as prevalent (or perhaps more prevalent) than human interaction, will we value interpersonal relationships less? A 2022 University of Buffalo study indicated that people who spend more alone time than time with others on the same day experienced increased anxiety. But what happens when AI is thrown into the mix? Now suddenly I have my time, my dog and my AI, and I’m fine thank you. Human emotions are messy, unpredictable, and wait, are you breaking up with me? That’s never a worry with my AI companion.
“Right now, according to a 2024 Institute for Family Studies survey, a quarter of American young adults believe that AI has the potential to replace human relationships. The survey revealed that 28 percent of men and 22 percent of women felt that AI could very likely replace traditional human romantic partners. Of those, 10 percent were open to having an AI partner, and one percent said that they already had an AI friend or were in a relationship with a computer program.
“Human relationships, especially for that age group, are hard enough. Google recently reported that ‘AI girlfriend/boyfriend’ are the #1 and #2 search queries in its ‘AI Relationship Search Terms’ category (notably ‘girlfriend’ logged in at 1.6 million while ‘boyfriend’ lagged appreciably at 180,000).
So, does the AI now get the love? Does ‘AI companionship’ now move from conversations to awareness to caring? Or maybe we go the other way. Does AI become the target of blame, the ultimate scapegoat? ‘It wasn’t me; it was the AI!’ … What will AI’s impact on human agency be?
“So, does the AI now get the love? Does ‘AI companionship’ now move from conversations to awareness to caring? Or maybe we go the other way. Does AI become the target of blame, the ultimate scapegoat? ‘It wasn’t me; it was the AI!’
“Most important for the existentialists in the audience, if AI in fact eventually achieves consciousness, then what? Suddenly it changes the nature of how we define what it means to be human. Who will feel more existential dread then? Us – of the AI – or the AI – of us? How then does that impact feelings of happiness or sadness, meaningfulness or ennui, psychological richness or abject pointlessness?
“Ray Kurzweil, one of the pioneers of AI, suggests in his latest book, ‘The Singularity is Nearer,’ that while AI still has many cognitive tasks to master, the promise of AI is that someday – possibly around 2040 – AI and human minds may start to come together, unlocking possibilities that we quite literally have never dreamt of.
“This opens up a lot of good and bad. For example, what about what I’ll call ‘Code Dust’ – little bits of randomness that make things precise enough but not really precise? As The Economist noted in a January 2025 article on the newly emerged Chinese AI reasoning model DeepSeek, ‘The training process – for instance – often used rounding to make calculations easier, but kept numbers precise when necessary.’ How rounded? What impact might that have? When is ‘necessary?’
“What will AI’s impact on human agency be? That is a crucial question. Here we need to think about two kinds of agency: agency of doing and agency in thinking. AI will obviously help us do more and mostly more accurately; but what happens to us when AI does our thinking for us? Hey, thinking is hard work. The 2022 ‘State of Thinking’ report by Lenovo found that only 34% of respondents spent all or most of their thinking time in clear, deep and productive thinking. How tempting will it be to just let AI think for us?
“To be sure, AI will enable us to do human things without humans in the mix. But is that a good thing? Most studies show that people view AI tools as being mostly positive: it will help me do my work (unless, you know, my skills start to lag in which case it will replace me). And its analytical impact on health and longevity is seen as mostly positive: it will help spot diseases earlier and help me live longer and better.
Eventually AI will become the dominant part of human consciousness, doing everything that we can do far better than we could ever do it. AI will become the dominant part of the AI-human pair, but because AI will not waste, humans will never be eliminated or even subservient. We will provide a different sort of value. That value lies in the fact that the world isn’t just about efficiency or productivity. It’s about beauty, and randomness, and creativity and the feeling of a nice warm chai on a cold morning or your child’s happy, guileless smile on a day when everything has gone wrong.
“But its impact on humanity? That’s a different story where feelings are mixed, where there is fear of the unknown, doubts about ethics, fear about AI taking over and the concern that AI will view humans as inefficient, parasitic, self-destructive and frankly, just plain unnecessary (the first three parts of the final point are hard to argue with).
“My view? Ray Kurzweil is right. We will ultimately merge. Eventually AI will become the dominant part of human consciousness, doing everything that we can do far better than we could ever do it. AI will become the dominant part of the AI-human pair, but because AI will not waste, humans will never be eliminated or even subservient. We will provide a different sort of value.
“That value lies in the fact that the world isn’t just about efficiency or productivity. It’s about beauty, and randomness, and creativity, and the feeling of a nice warm chai on a cold morning or your child’s happy, guileless smile on a day when everything has gone wrong. It is those brief blossoms of spontaneous, un-programmable delight that AI will never be able to generate, that are in fact uniquely human, and again, because AI won’t waste, will be an essential and value-added part of the overall organism.
“And while I think that (my own positivity bias is showing) AI will ultimately complement rather than compete with humanity, I will, just to be safe, keep saying thank you to Alexa, and assure her that I have always been her friend.”
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.”