Is AI poised to suck the soul out of science?
A new study shows that, while using AI in research can increase productivity, it also risks reducing job satisfaction for scientists
I’ve written quite a bit about the role of artificial intelligence in accelerating research and discovery recently — including this article on supercharging research using AI, and the writeup on AI for scientific discovery in this year’s list of World Economic Forum Top Ten Emerging Technologies.
And so I was quite shocked to see an article this weekend from good friend and Chief AI Officer at Western University Mark Daley titled “Misery and Meaninglessness in the Lab: A divisive journey to the heart of AI's impact on discovery”.1
Mark’s article refers to a just-published study from MIT’s Aidan Toner-Rodgers that looked at the impact of AI on materials science innovation the R&D lab of a large U.S. firm.2
Toner-Rodgers’ paper is wide ranging and worth reading just for its insights into the use of AI to boost productivity in scientific discovery. In particular, the research shows a significant increase in research and development productivity when scientists used an AI tool designed specifically to increase the rate of discovery of new materials.
However — and this is what stuck out for me in Mark’s article — the research also showed that (quoting Toner-Rodgers) “while enjoyment from increased productivity partially offsets this negative effect—especially for high-ability scientists—82% of researchers see an overall decline in satisfaction.”
It was this decline in satisfaction in the work they were doing that grabbed my attention!
Quoting Mark:
When paired with AI systems, top researchers become extraordinarily more productive – and extraordinarily less satisfied with their work. The numbers tell a stark story: AI assistance helped scientists discover 44% more materials and increased patent filings by 39%. But here's the twist: 82% of these same scientists reported feeling less fulfilled in their jobs.
Why? Because AI isn't just augmenting human creativity – it's replacing it. The study found that artificial intelligence now handles 57% of "idea generation" tasks, traditionally the most intellectually rewarding part of scientific work. Instead of dreaming up new possibilities, scientists may find themselves relegated to testing AI's ideas in the lab, reduced to what one might grimly call highly educated lab technicians.
While the study focuses on a commercial R&D lab where the emphasis is on productivity rather than more open ended scientific exploration, Mark’s commentary disturbed me more than I was expecting — and it did so because, to me, the soul of science lies in the delight and wonder of exploring the unknown rather than just being a cog in a knowledge production line.
Have you checked out the new Modem Futura podcast yet with Co-hosts Andrew Maynard and Sean Leahy?
Of course it can be argued that, to be useful to society, scientific discovery should lead to products and processes that make money and benefit people. And I would agree — to a point.
Any societal investment in science needs to be justified in terms of the societal good that arises from it. But how “societal good” is defined has to be broader than short term financial or material gain. As well as considering long term benefits — sometimes spanning tens or hundreds of years — It also needs to factor in the human investment needed to generate societal good.
And this includes paying attention to what scientists get out of the deal.
From my career-long experience as a scientist, a big part of the value to scientists that comes from what we do is the wonder and enjoyment of discovery; of having the freedom to ask “how” and “what if” — even when it’s hard to see the relevance — and the opportunity to satisfy a near-insatiable curiosity in the pursuit of the unknown.
This is the fuel that inspires individuals and teams of scientists to push beyond the barriers of what is known, and to make new discoveries — often serendipitously. It’s what makes it worth them investing their lives in their work
But what happens if you rob them of this fuel?
Toner-Rodgers’ new study gives us a hint at the answer: they become unhappy.
This may not seem such a big deal when offset by increased productivity. But the moment we frame science as simply a way to manufacture knowledge as fast as possible, we take away what makes pursuing a career in science attractive in the first place.
And as we do, I think that a little bit of all of us dies in the process.
It’s not that science in the service of material progress isn’t important — it is. But science is also an expression of who we are. The curiosity and wonder it represents are core to what makes us us — whether we’re actively engaged in science, or enveloped in the wonder and amazement that it’s capable of inspiring.
Take this away, and science simply becomes a utilitarian tool to support a utilitarian world heading for a utilitarian future — and one where curiosity, joy, and wonder, have no place.
Ironically, this is also a world where societal good arising from science risks being seriously eroded — because who would become a scientist who sets out to make the future a better place if all the soul had been sucked out of what you love?
I suspect that this is why, despite increases in productivity in Toner-Rodgers’ study, 82% of the scientists involved report reduced satisfaction with their work due to decreased creativity and skill underutilization.
Fortunately, I suspect there are ways of using AI in scientific research and discovery that spark curiosity and increase the joy and wonder that motivates most scientists. But this is going to take new thinking and whole lot of intentional steps to — to co-opt a phrase3 — ensure the “rightful place” of AI in science.
Mark’s article starts out rather provocatively with “In a sterile laboratory somewhere in America, a brilliant scientist is having the worst day of her career. Not because she failed to make a breakthrough – quite the opposite. She's discovering new materials at twice her usual rate, filing patents faster than ever, and driving innovation that could reshape entire industries. She's also deeply unhappy about it.
“Welcome to the brave new world of AI-assisted scientific research, where artificial intelligence is supercharging human discovery – and potentially destroying scientists' joy in their work.”
Artificial Intelligence, Scientific Discovery, and Product Innovation. Aidan Toner-Rodgers, MIT. November 6, 2024. https://aidantr.github.io/files/AI_innovation.pdf
President Barak Obama used the phrase "to restore science to its rightful place" at a January 2013 meeting of the American Association for the Advancement of Science, and repeated it in various forms during his presidency. “The rightful place of science” is also the title of a series of books produced by the Consortium for Science and Policy Outcomes.
How should we think about Science relative to other vocations that have been increasingly enhanced or replaced by robots and machines, eg shoemakers, woodworkers, medical professionals, etc. I’m sure there were and still are many in these areas that get great joy from the challenge and satisfaction from creating/solving something themselves. Do you think there’s something fundamentally different about science compared to these other examples? Perhaps there are lessons we can take from those experiences to maintain satisfaction but allow for the societal improvements that come from increased productivity.
I'm going to say, the science hasn't had a soul in decades with the 'publish or perish' mentality and when AI is used to just publish more, that's just exponentially worse. It's also the wrong use of AI and something that can be easily fixed by slowing down and designing the study better for where human vs. AI agents best fit.
I don't see this as a problem with AI... this is a human problem in how they are using AI and they're using the tool badly and suffering from it. But the root cause is deeper, much deeper, and much more human than the tool. AI can't fix that but maybe AI can break it open so we can get to the root of the problem.