To Hell With the Ant. Long Live the Grasshopper
It's not AI or other tech that makes people less conscientious. It's life in 2025.
Consider Aesop’s fable of the ant and the grasshopper. That’s the one where the ants drudge away during the summer, piling up food, while the grasshopper frolics and plays music. When winter comes he has nothing to eat, and the ants say, dude, told you so. (I guess the story dates from before commerce, since no one suggests they trade food for music.)
The ants have two traits in this story that are often conflated. First, they are, in the words of this piece (paywalled) by Financial Times’ John Burn-Murdoch, “dependable and disciplined.” Second, they (his words again) “put long-term well being ahead of short-term kicks.” These aren’t exactly the same thing (more on that in a minute) but they’re usually treated as one by the $46 billion “self-improvement” industry and by bosses, media pundits, academics and other Serious People who’ve selected themselves to tell the rest of us what to do. They call it conscientiousness. People who score high on this trait, according to Big Self-Improvement and a fair amount of research, do better in life, and make society better.
Conscientiousness is one of the “Big Five” personality traits that most psychologists believe are consistent across times and places. (The others are “Openness to Experience,” Extraversion, Agreeableness and Neuroticism.) In “Big Five” theory we all fall somewhere on a high-to-low spectrum for each trait. For conscientiousness, that goes from low (grasshopper) to high (ants). (You get your score from assessing yourself on survey questions like “I seek variety over routine” and “I like to tidy up.”)
Earlier this month, Burn-Murdoch’s article got a lot attention because it claims that people have become less conscientious over the past decade, and that digital technology is likely one of the reasons. Pundits and commentators rushed to deplore this development, and commenters tsk-tsked in great numbers.
I want to unpack this because I think a lot of the tsking is misplaced. And relevant to the weird AI-adoption moment we are in.
I read Burn-Murdoch’s claim as four parts. First, that conscientiousness across society can be measured; second, that it’s really steeply declining; and third, that this decline can be explained by personality psychology (ie, people’s individual experiences and choices, rather than politics or the economy); and, fourth, that this is to be deplored.
Let’s take a look at each of those claims.
Can Conscientiousness Be Measured?
For this, we need to define “conscientiousness.”
Most definitions, like Burn-Murdoch in his piece, combine two different traits. One is doing what you’re obliged to do (because you said you would, because it’s your job, because that’s what the rules say, etc). The other is discounting the present in favor of the future (save money now to have it in 2050, work hard for your Master’s degree to get a good job, etc.) These two qualities intersect in obvious ways, but they aren’t exactly the same.
Doing what you are obliged to do, following the rules, keeping your word, has been valued since ancient times in most every culture. No one likes a cheater. (Though a lot enjoy stories of tricksters who evade the rules.) But the self-mastery and honor of our ancestors differs from modern social science’s conscientiousness in an important way.
Conscientiousness in the Abstract
“Conscientiousness” in personality psychology is a trait: You have a certain amount of it, consistently, no matter where and when you are. Psychology textbooks define conscientiousness as, for example, “the propensity to follow socially prescribed norms for impulse control, to be goal-directed, planful, able to delay gratification, and to follow norms and rules.” Note what is missing: What norms? Which rules? Plans for what? Are we planting trees in the park or opening a concentration camp?
The conscientiousness of ancient and traditional societies isn’t like that at all. It was rooted in particular places and particular people. You kept your word to people you’d known all your life, people who, if you weren’t reliable, would suffer. With others, you might not.
This wasn’t an abstract, Enlightenment virtue, given to all. It was for “us,” and not for “them.” As the historian Patrick Joyce writes in Remembering Peasants, this is alien to our “modern” ideas of conduct. (He quotes a political scientist deploring the “amoral familism” of mid-20th century Italian peasants, whose norms and rules extended, he said, only to relatives. Any value judgment there? Well, the book is called The Moral Basis of a Backward Society.)
And, sure, selective conscientiousness can be bad for society. Why should you treat me differently because I’ve just arrived in town? Not fair!
But limits to conscientiousness can guide resistance to oppression. Peasants, army conscripts, serfs and slaves have refused to be conscientious about work assigned by those who oppressed them. The labor Polish peasants were forced to do for their overlords, Joyce writes, “was often deliberately done badly and with contempt – as a sort of peasant revenge.”
The labor Polish peasants were forced to do for their overlords “was often deliberately done badly and with contempt – as a sort of peasant revenge.”
Enslaved people, too, didn’t extend conscientiousness to their oppressors. Here’s Constantin Volney, the French revolutionary, describing slaves planting peas during his visit to a plantation in the south of the United States in the 1790s.1
The master took a whip to frighten them, and soon ensued a comic scene. Placed in the middle of the gang, he menaced, and turned far and wide (on all sides) turning around. Now, as he turned his face, one by one, the blacks changed attitude: those whom he looked at directly worked the best, those whom he half saw worked least, and those he didn’t see at all, ceased working altogether; and if he made an about-face, the hoe was raised to view, but otherwise slept behind his back.
What is captured in psychologists’ measure of conscientiousness is a generic self-assessment, without a context. It doesn’t allow for the distinction these forced laborers made. In omitting that, it can describe as a deficiency what should better be seen as resistance. Not organized. Not principled or even effective, either. But better than nothing.
Is Conscientiousness Really Declining Dramatically?
Still, there is something real being measured here – people’s assessment of their own personalities. Now let’s look at what Burn-Murdoch made of it.
His data comes from the Understanding America Survey – a representative sample of thousands of U.S. households, whose members respond to Internet surveys designed by social scientists. Comparing 2022 data with 2014, Burn-Murdoch sees a marked decline in conscientiousness scores. One that, he says, is especially steep in younger people. A number of statisticians quickly disagreed, saying he’s greatly exaggerating a small trend. Yesterday’s account by Elizabeth Nolan Brown sums up their reasons. She makes a good case that things aren’t going as dramatically downhill as Burn-Murdoch wrote.
Can This Supposed Decline Be Explained By Personality Traits?
But let’s stipulate – just for a moment! – that even a small decline could be bad news for society, because being an ant is so much better than being a grasshopper. That lets us consider the third aspect I mentioned: That the cause of this decline is best understood by looking at individual psychology. “The sheer convenience of the online world makes real-life commitments feel messy and effortful,” Burn-Murdoch writes. “And the rise of time spent online and the attendant decline in face-to-face interactions enable behaviors such as ‘ghosting.’ ”
Seems plausible. For example, it’s often said lately that AI companions (as in Replika.ai and other services) could make people less inclined to socialize with each other, because the ersatz humans are much more agreeable to hang with. However, I’m struck at how many of the people who have said this are speculating from a distance. When I look at research on actual users, it suggests a more ambiguous effect, with at least some people improving their social lives after using AI “friends.”
If tech effects aren’t so clear-cut, maybe we can look to explain a decline in conscientiousness in a systemic way. We might, for example, consider that the job market is shrinking. Employers sense they have the upper hand, and are making a show of cracking the whip.
Capitalist mythology, cited in this CEO’s memo, is that if you don’t like that, you can find another job. Capitalist reality is that many people will, for one reason or another, be stuck under the whip. So maybe they will be less conscientious on the job. Do they have defective personalities, deformed by technology? It might be more accurate, or at least more historically informed (remember the Polish peasants and enslaved Black Americans), to blame the economy that has trapped them under ruthless executives blabbering about alignment and not believing in work-life balance.
A pleasing irony about this kind of passive resistance is that it can make use of generative AI – the same technology that bosses are using to replace and frighten the workforce. Maybe some day soon AI will replace you entirely, cutting costs for your employer. In the meantime, though, as this Apple ad reminds us, AI gives workers ways to slack off a bit.
Of course, life is better for individuals and society if everyone is dedicated to a job and loves putting in 110 percent. Who wants to drag their feet and malinger when they could be full of enthusiasm and shared purpose? Life would also be better if there were no more wars and climate change magically reversed.
Resistance is a part of life in the real world. The fix won’t be found in blaming individuals and the technology they use. It will be in creating an economy in which fewer people feel trapped and desperate.
Is a Decline in Conscientiousness So Bad, Under the Circumstances?
Which brings me to the fourth aspect of Burn-Murdoch’s argument, which is conventional wisdom: That a decline in conscientiousness is to be deplored and, if possible, reversed.
I don’t think a democratic society should, or could, eliminate the quiet resistance of people who labor under duress. It’s an ancient aspect of any society that isn’t an egalitarian band of hunter-gatherers. “When the great lord passes, the wise peasant bows deeply and silently farts,” is ancient wisdom. Sure, it would be better if everyone joined a union and rallied to change society, but you have to start somewhere. Refusing the Kool-Aid of alignment with Corporate can be that start.
All of the above has been about one aspect of conscientiousness (being reliable). But I’m also skeptical about the supposed badness of a decline in the concept’s other aspect – future orientedness. Should we worry that people don’t sacrifice today for tomorrow as much as they used to?
When Today Matters More Than Some Imagined Tomorrow
Being future-oriented makes sense in a world of fabled ants and grasshoppers, where the past is a perfect guide, and everyone knows what the future will bring. The future, though, has never been so cooperative. It is full of surprises, which means our calculations about today vs. tomorrow are clouded in uncertainty. This does not mean we shouldn’t look ahead and make what plans we can. It does mean that we should not, and cannot, say for sure that this deed done today will lead to that outcome tomorrow.
This is always true, but it is evident to more people in times like ours, of technological, economic and political change. Remember when future-minded, practical students majored in Computer Science? Lots of jobs there! Except now, times have changed, and Art History grads are more likely to get a job than CS majors. Should my teen-age son work hard to get into an elite college in order to get into an elite profession? This was the meritocratic mantra of my social class as a kid, and prevailed for decades in neighborhoods like mine. But lately? What elite jobs will be left after 20 more years of AI infusion? What will the rewards of joining the meritocratic tribe be, after more years of populism? He’d be better off studying what pleases him, and taking his chances.
So when people report that they are “less likely to make and deliver on commitments,” as Burn-Murdoch writes, they may be reacting rationally to their lived experience, in an environment of rapid technological and economic change.
New Laws Protect the Vulnerable from AI
Most people seem to be able to handle the pretend-play of chatbot interaction. They know ChatGPT or Claude or Replika.ai is just software and hardware, just as they know that Superman is actually an actor, even if he sparks real emotions at the movies.
But some aren’t able to maintain this double vision. Bot interactions send them spiraling into delusions. As evidence mounts that this is a serious problem, governments are looking for ways to protect the vulnerable.
So, in New York State, AI chatbots will soon be legally required to remind users, at least every three hours, that the bot they’re chatting with is not a person. The new law takes effect November 5. California looks likely to adopt the same three-hour requirement soon.
How Much Specialized AI Do We Need?
There are a lot of startups out there offering human-like AI for therapy, friendship, coaching, “digital necromancy” (AI versions of the dead) and, of course, sexual partners. One obstacle they all face is the fact that plain-vanilla general-purpose AI can, with a little ingenuity, perform a lot of the same services. Case in point: Apparently it’s pretty easy to get ChatGPT to help you masturbate.
Less Pain, Less Gain
As a lot of us have noticed, it’s easy – very easy – to get genAI to gather up a lot of specific information in response to a question. Too easy, say these researchers. The lack of effort required to get information from AI, their experiments found, “leads users to develop shallower knowledge.”
Does Tech Use Help Fend Off Dementia?
As time passes, it’s becoming clear that “the same” genAI can have very different effects on one person than on another. As I mentioned above, the impact of a “companion” AI on a typical person versus a person with mental-health issues is an example. Another might be age-related. For all that tech use is associated with less learning and engagement in kids, this analysis of 57 prior studies on digital tech use found that, in people older than 50, use of these tools is associated with lower risk of dementia and slower cognitive decline.
For a long time this “master” was thought to be Thomas Jefferson. Historians have shown this isn’t so.↩︎
I think life is about balance, not binaries. We need to be able to look to the future to be able to build a better tomorrow, but we also need to be in the moment to make sure that we are not ignoring the injustices of today.