We’re talking a lot these days about using AI to automate drudgery. But I’m starting to think there is a hidden cost to letting it eat all the low-hanging fruit.
On this Oceans trip to Sri Lanka, I led an internal training on proactivity and how to bring value to a client. And I started with a little roleplay: I’m in the kitchen cooking dinner and you come to help. What would you do?
There were a variety of available options (from “taking out the chicken that is on fire” to “doing my taxes that aren’t due for nine months”) but what was meant to be the Low Importance / Low Urgency option was cleaning a scuff mark on the wall that has been there forever.
One by one, I brought people up and asked them to pick the task they would do to bring value to me, then explain their choice. And I was shocked when one of our Divers picked the scuff on the wall. But his reasoning is thought provoking: “It is the easiest thing to fix immediately, so I can do that and create some momentum to do the next thing.”
I’m not saying that’s the right answer (take out the burning chicken, please and thank you) but it is very much a technique that applied behavioral scientists recommend; small, easy successes can help create a snowball effect that allows us to take on bigger things. And it respects human dignity; we’re not always in a place where we can tackle something urgent and important.
Both Steve Jobs and Tim Cook were famous for starting their days by responding to customer emails. From the perspective of the CEO, those individual concerns were neither important nor urgent, and they are exactly the kind of messages that AI is being used to automate.
Both executives swore by the practice and the degree to which it kept them grounded and close to what really matters. It is easy, whether they are done by someone far down the org chart or automated with AI, to forget all of the small things that make it possible to do big things.
And though neither mentioned it, I think equally important is the snowball effect: starting your day by using the power of your office to fix the problems of end users is a pretty good way to meet the urgent, important work with a full head of steam. If we leave ourselves only the fruit at the top of the tree, each day will start with a longer and longer ladder that becomes increasingly daunting to climb.
I’m not suggesting that we pull the emergency break on the automation train or that the CEOs of the world should be answering all the customer service questions. But whatever we automate, we should also allow for manual contribution. I have a dishwasher but would mourn a world where I was compelled to always use it. Because sometimes I want to do them myself, to look out my kitchen window and marvel at the world, with my hands in warm soapy water and the immediate fruits of my labor visible: clean glasses, like a miracle, a product of my own direct effort.
Britt Titus argues that behavioral science is poorly equipped for conflict settings because most of its theories, methods, and evidence come from stable environments. With billions of people living in conflict-affected areas, this gap is not just academic—it has real consequences. Research in these contexts is difficult, but the bigger issue is that the field often prioritizes what is easy to measure over what truly matters, leading to shallow understanding and limited impact.
She emphasizes that conflict doesn’t just intensify existing behaviors—it fundamentally reshapes how people think about risk, trust, time, and cooperation. What might appear as “bias” in standard behavioral terms can actually be rational responses to extreme uncertainty. For example, prioritizing immediate rewards over future gains may reflect a realistic assessment of survival, not a cognitive error. Misinterpreting these behaviors risks designing interventions that are ineffective or even harmful.
Titus also argues that simply adapting interventions from other contexts is insufficient. True effectiveness requires co-designing solutions with local communities from the start, ensuring they reflect local values, needs, and realities. When this approach is taken, interventions become more relevant and are more likely to be adopted. In contrast, lightly “localized” programs often fail because they remain rooted in assumptions from entirely different environments.
Another key challenge is scaling and measurement. Many promising interventions remain small-scale pilots, with little understanding of how to expand them to broader populations. At the same time, the field tends to measure attitudes—like trust or empathy—rather than actual behaviors, even though real-world outcomes depend on what people do, not what they say.
Overall, Titus argues for a shift in approach: start with communities rather than predefined problems, co-design interventions with local actors, test assumptions early, and design for scale from the beginning. Most importantly, behavioral science must recognize that conflict settings are not just “harder versions” of stable ones—they require fundamentally different ways of understanding and intervening in human behavior.
👉 Read the full piece here
You’re not “just” a behavioral scientist — you’re someone who knows human behavior is a competitive advantage. Here are 5 companies hiring people like you this week 👇
Behavioral Research Scientist - Peraton
People Research Scientist - Anthropic
Senior Researcher, Human and Community Development - American Institutes for Research
Sr Director, Audience Insights - Microsoft
Head of Insights - Pentasia
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.