New research suggests using GenAI depletes our critical thinking skills
An interesting new study done by Microsoft and Carnegie Mellon University (article behind paywall unfortunately) has found that the use of generative AI means that our cognitive and critical thinking skills diminish. The key conclusion of this study to my mind was that as workers trusted AI more they stopped looking for the errors and mistakes that AI was making; in doing so, it turned out they were also losing their ability to think critically. That is my two-fer: content from AI is still far from being trustworthy - unreliably unreliable - and yet soon we won’t have the critical skills to even realize we’re reading rubbish. Great that’s all we need, but we should not be surprised.
“[A] key irony of automation is that by mechanizing routine tasks and leaving exception-handling to the human user, you deprive the user of the routine opportunities to practice their judgement and strengthen their cognitive musculature, leaving them atrophied and unprepared when the exceptions do arise,” the researchers wrote. This finding generalizes across many areas of human endeavour. For instance, it is one of the major ideas behind the Free Range Kids movement. If adults completely control and dominate the interaction of children, we should not be surprised that they have little independent thinking or autonomy.
“The data shows a shift in cognitive effort as knowledge workers increasingly move from task execution to oversight when using GenAI”. This reminds me of what happens when you go into a retail store - which are usually staffed by people under 40 - and suddenly the automated VISA approval process goes down. They want to close the store. Wrong! You can actually still use a telephone to get the visa approved. We have just forgotten how to. (Dead giveaway that I’m older because I’m actually in a store and not in some virtual world at home where if I click on something the influencer told me me to, a product will show up at my house in 2 to 5 days.)
“The researchers also found that “users with access to GenAI tools produce a less diverse set of outcomes for the same task, compared to those without.” Another thing we definitely don’t need. I remember in the early 1990s as the internet emerged (who calls it cyberspace or the information superhighway anymore) it was sold as a influential and creative force opening up the amazing variety and richness of human thought and interests. Blah blah blah. Instead, I don’t think we’ve seen such conformity since churches ruled the roost. I call it the clip and paste era. I’m surprised when I actually something new, fresh, and vibrant.
The article concludes somewhat positively by stating, through a complete set of cherry picked examples (Socrates the great orator was against writing), that society always goes through this process when major new changes happen. A few little blips in the road then everything is happy and merry on the other side. I’m not sure that’s ever true. We get Pros and we get Cons. Instead we should try not to get conned into blindly accepting the Cons.
A Pre-owned WORD SALAD
I am not actually recycling one of my previous articles, I’m just trying to introduce this week’s word with a bit of a trick. Preowned is one of those euphemisms where if you use the previous word - used (worn, old) - people’s eyes dart and their breathing speeds up, especially if they are vendors. It is like you’ve done something quite unsavoury, like telling a vulgar joke, not just saying the truth. The expression’s descent to today’s ludicrous term started with secondhand, which is also a bit silly. I think of secondhand as my non-dominant right one which is pretty useless. Then we got the direct precursor term: Previously owned. That is technically correct, but is starting down the path to preowned’s meaninglessness.
About a year ago I listened to a fascinating podcast by John McWhorter, who is a famous American linguist. In it, he pointed out how many terms start by having neutral or even positive connotations, and they then over time they become pejorative due to societal attitudes or overuse. He used examples, such as racial terminology, to support his thesis. I get that. So, for example, one of the current politically approved terms - People of Colour- likely will go through the same downward spiral. We just don’t know when nor do we even have any clue of what will replace it.
But what did poor old “used” do to deserve this treatment? I submit that there’s another reason that words go through this wheel, becoming discarded or pejorative. That is to avoid the reality of the expression. Somehow if you put the word preowned in front of car it is suddenly new, especially if it’s been reconditioned or refurbished. We love to fool ourselves. I cannot keep up with what the current passive euphemism is for death and dying; maybe poor old Dad has left the room.
John McWhorter also writes about the change process for languages, how nuances and new meanings emerge all the time. My second proposition is that this process of change is happening more rapidly, caused by technology. Everyone is spending many hours each day ingesting an avalanche of little (often nearly meaningless ) data points, whether textual, oral, or primarily visual. On top of this is our fascination on every platform with “Trending”. Then with our likes and reposts we create memetic phenomena. This is the acceleration by which new words and expressions are quickly normalized. Yes that last word was meant to be ironic.
Thanks once again as always for reading. Appreciate the comments as they keep me focused and motivated
Critical thinking skills have been going downhill for a while now. AI is just going to accelerate that in some ways. I honestly don't know how people are gonna learn to read and write in the future. It's gonna be a completely different way of learning.
As we let AI take over more tasks, our brains kind of go on autopilot, trusting technology without questioning it. It's like a slow fade of our cognitive muscle, and before we know it, we won’t even spot the flaws in what AI is churning out. Admittedly, this happens to me now ugh. The only difference is I am more self-aware. We were promised a world of infinite variety when the internet emerged, but it seems like everything’s become a recycled version of the same few ideas. AI tools, like many tech innovations, can create more uniformity in our work, not necessarily more creativity.
Thank you for the conversation, David.
Happy Friday.
Happy Valentine's Day!