|
Humanity has been worrying about the potential cognitive downsides of technology for a long, long time. As Socrates is reported to have said about writing: “It will introduce forgetfulness into the soul of those who learn it: they will not practice using their memory because they will put their trust in [it], which is external and depends on signs that belong to others, instead of trying to remember from the inside, completely on their own. You have not discovered a potion for remembering, but for reminding; you provide your students with the appearance of wisdom, not with its reality.”
It is easy, in hindsight, to roll our eyes at the suggestion of cognitive harms associated with learning to read and write. And yet some of Socrates’s fears did come to pass: you and I are almost certainly less skilled at remembering tens of thousands of lines of epic verse than someone who was raised in a predominantly oral tradition.
For people who use AI to help them get work done in their professional and personal lives, a similar transition seems to be occurring. As we report this week, the use of a cognitive enhancement does tend to diminish our ability to function unaided.
Ask people to write an essay using an AI assistant, for example, and their ability to recall the contents of that essay is poorer than it is among those who wrote theirs by hand. Ask them to come up with innovations using an AI chatbot, and they get worse at generating novel ideas without it than those who were never given access to the chatbot in the first place.
The effect seems to linger. A study of British AI users found that those who used and trusted AI most scored lower across the board on a critical-thinking assessment—although whether the two trends were causally linked (and, if so, in which direction) or simply correlated remains unclear.
The core question, for critics and supporters of AI alike, is whether this latest technology is distinct from those that have come before. For the ancient Greeks, rote memorisation was an important part of what it meant to be human. Now, it’s a party trick. Does the same fate await brainstorming?
I’m always wary of claiming to live in exceptional times, but the rise of AI does feel meaningfully different to previous technological advances. If a student uses a calculator in a maths test, they still have to understand the principles at hand. If they’re allowed reference books, they still need to make a coherent argument. If they hand off the writing of an essay to an AI assistant, however, what’s left for them to do? Will AI be the latest cognitive enhancement or the first fully fledged cognitive replacement?
Such debates are worth having, but these issues remain hypothetical. For now, AI is good, but not great. Using AI efficiently remains a skill, just like using a calculator or, indeed, pen and paper. Learning how to use the technology as an aid, rather than a substitute for effort, may develop new skills even as it lets old, obsolete ones wither.
Elsewhere this week:
Thank you for reading. How do you think AI will affect creativity and learning in the human mind? Send your thoughts to
sciencenewsletter@economist.com.
If you’re struggling to think of an answer, a chatbot might be able to help you. |
