As I’ve been experimenting with AI tools like ChatGPT, Claude, Lovable.dev, Fireflies.ai, and other emerging AI agents, I was reminded of a TED Talk I watched in 2012 by behavioral economist Dan Ariely titled “What Makes Us Feel Good About Our Work?”
Using these tools has brought moments of joy, especially when AI takes over repetitive tasks. But it’s also triggered feelings of frustration and even loss, especially when AI encroaches on areas I still want to own. This tension made me reflect on how important it is that we intentionally design this new AI-powered world to support human achievement rather than undermine it.
What really motivates us?
In his TED Talk, Ariely explores what drives us at work. He argues that meaning, recognition, and a sense of purpose often outweigh monetary rewards in motivating us.
He distinguishes between two kinds of work:
- Meaningful, challenging work: These are the tasks where we demonstrate mastery, overcome obstacles, and feel progress. This type of work energizes us and builds pride. Think of training for and completing a marathon, or delivering a high-stakes presentation that earns praise.
- Sisyphus-like tasks: Repetitive, menial tasks can drain us—leaving us demoralized. Examples include taking notes during meetings, reading through massive legal documents, or slogging through administrative work.
AI is remarkably capable of handling both types of tasks, but whether it should is the more critical question. Sometimes, handing everything off to AI can do more harm than good.
Like letting a toddler dress themselves
This reminds me of when toddlers start expressing independence, wanting to pick out and put on their own clothes. It takes patience for a parent to step back and let them struggle through it, knowing they might be late for work. But that struggle is essential. It’s how kids build confidence and capability.
I felt a similar dynamic when I hired a service to cook meals for my family.
From outsourcing to ownership in the kitchen
Both my husband and I work full-time, and neither of us came into adulthood with strong culinary skills. When our kids were young, we wanted to break out of the mac-and-cheese cycle and offer a variety of healthy meals. But the thing we dreaded most? Grocery shopping. We’d get overwhelmed, buy random ingredients, and watch them wilt in the fridge.
So we hired a local service called The Food Fairy. It felt life-changing at first. They planned our meals, did the shopping, cooked in our home, and left delicious, gourmet meals ready to heat up. But over time, the drawbacks surfaced: the kids grew picky, the cost added up, and on non-service days, we still didn’t know how to cook.
A friend introduced us to HelloFresh, and it turned out to be a perfect in-between solution:
- We select meals as a family, giving us variety and agency.
- Ingredients arrive at our door—no grocery stress.
- Everything is portioned perfectly, so no waste.
- Step-by-step instructions build our confidence in the kitchen.
- It’s more affordable and sustainable for us.
- And—most importantly—we’re learning.
HelloFresh removed the worst part (shopping and waste) while keeping the best part: learning by doing. When we make a meal and our kids eat it without complaint? That’s a win. That’s pride.
Some AI support systems should be like HelloFresh
There’s a lesson here for how we build AI agents. For skill-building tasks, I don’t want AI to take over. I want it to be like a coach: guiding, supporting, and challenging me to improve.
When AI creates for me, without my input, I feel disconnected. Even resentful. It’s like being sidelined from my own creativity.
Here are some exciting AI Agents that are applying a coaching mentality to improve skill development
- Mercor: Provides AI-generated mock interviews with real feedback for prospective job seekers to practice their interviewing skills
- Pitch Guide: Provides entrepreneurs specific feedback on their Pitch Decks and recommendations for improvement
But for sisyphus-like tasks? AI should take it all
One of the first AI tools I integrated into my workflow was Fireflies.ai, a meeting note-taker. Before that, I would juggle scribbling notes while trying to facilitate meaningful conversations. Often, I missed things or couldn’t decipher my own handwriting later.
Fireflies now records and transcribes my video calls, allowing me to stay fully present in the conversation. I still jot down notes during meetings, but I no longer worry about missing key details.
After a meeting, I walk away with:
- A full transcript for reference
- A well-organized summary I can edit and share
- Action items clearly assigned
I don’t need this level of support for every call, but for prospective client meetings, it’s a game changer. It lets me follow up quickly, stay professional, and focus on what matters most.
Designing work that works for humans
Dan Ariely’s insights feel more urgent than ever. His research offers a powerful blueprint for how we can build our AI-augmented future.
- Use AI to challenge, guide, and grow our mastery but not replace it.
- Let AI relieve us of repetitive tasks, freeing up energy for connection, focus, and deeper work.
AI can be more than a tool. It can be a partner. But only if we design it with intention, humanity, and meaning at its core.
Looking for more AI resources and ideas? With the success of our AI in Focus learning series, we are continuing to host weekly live streams. Our initial sessions are hosted on our website, and as we produce more you can find them on our YouTube channel. Follow us on social media to be notified of upcoming livestreams.
If you have an idea for a way AI could be tapped to tackle a challenge or take advantage of an opportunity, we would love to discuss further.