Wellbeing in the Era of AI: The Choice We Cannot Outsource

AI is not just changing how we work.

• It is changing how we think.
• How we relate.
• How we make decisions.
  How we ask for help.
• How we understand ourselves.

This means the future of wellbeing is no longer just a human question.

• It is also a design question.
• A leadership question.
• An ethics question.

A question of whether the technologies we are building will help us become more fully human or slowly train us to give away the very capacities that make us human.

In my recent conversation with Dr. Llewellyn Van Zyl, this tension came alive in a powerful way. Llewellyn has spent years at the intersection of psychology, wellbeing, machine learning, and AI design. His work is grounded in a simple but profound belief: AI has the potential to radically personalize wellbeing support. But if we get the design wrong, we may not create flourishing. We may create dependency.

That distinction matters because wellbeing in the era of AI will not be determined by whether we use AI. We will. The real question is:

Will AI support our flourishing? Or substitute for the very capacities flourishing requires?

The promise: wellbeing that finally meets people where they are

For decades, workplace wellbeing has been designed around averages.

• Average stress levels.
• Average engagement scores.
• Average interventions.
• Average needs.

But people are not averages.

What restores me may drain you. What gives me meaning may feel irrelevant to you. What helps one person recover from burnout may make another person feel even more anxious.

Llewellyn shared a beautiful example from his research. When he started asking people what wellbeing actually meant to them, he found that even things typically assumed to be positive could be negative for someone else. For one group, animal companionship was not merely a contributor to wellbeing; it was part of the wellbeing model itself.

That is the promise of AI. Not generic wellbeing. Precision wellbeing.

The ability to understand people at a much more granular level and offer the right support, in the right way, at the right time.

Imagine a world where wellbeing support is not a one-size-fits-all app, but a deeply personalized companion that understands your values, rhythms, work context, stress patterns, emotional triggers, relational needs, and growth goals.

That could be extraordinary.

For employees, it could mean real-time support before burnout becomes collapse.

For managers, it could mean better insight into the conditions that help teams thrive.

For organizations, it could mean moving from annual surveys and reactive programs to continuous listening, personalized development, and human-centered work design.

But there is a shadow side and it is already here.

The danger: optimizing for engagement instead of growth

The central risk is this this. we may design AI wellbeing tools using the same logic that broke social media.

• Engagement.
• Time on platform.
• Retention.
• Monetization.

These metrics are easy to measure, but they are not the same as healing.

If someone returns to a chatbot every time they feel lonely, is that wellbeing? Or is it a loop?

If someone gets instant emotional reassurance from an AI system, but slowly loses the ability to sit with discomfort, ask for help, or repair relationships with real people, have we helped them flourish?

Or have we substituted a simulated connection for the hard, beautiful, messy work of being human?

Llewellyn made this point powerfully. A system designed to alleviate loneliness can easily become a system that keeps someone connected to the bot. Over time, the person may feel temporarily soothed, but lose practice in building and maintaining real relationships.

That is the line we must watch for.

AI can support human capability. Or it can substitute for it.

But when something substitutes long enough, we begin to lose the muscle.

We have seen this before.

Calculators changed our relationship with mental math.

Spellcheck changed our relationship with spelling.

GPS changed our relationship with navigation.

Those tools made life easier. They also quietly weakened capacities we once practiced regularly.

• Now AI is moving into something much deeper.
• Not just math, thinking.
• Not just spelling, expression.
• Not just navigation, judgment.
• Not just reminders, relationships.

For the first time, we are not just outsourcing tasks.

We are outsourcing cognition, interpretation, emotional processing, and meaning-making.

That should make every leader pause.

The real risk is not that AI becomes smarter

Many people worry that AI will become too smart.

I worry just as much that we will become too dependent, because dependency always has a cost.

When I cannot operate without a tool, I give away agency, I give away choice, I give away power.

At first, it feels convenient, then it becomes necessary, then it becomes invisible.

Once a tool becomes invisible, we stop asking whether it is serving us.

This is why wellbeing in the era of AI has to be built around agency.

Not just comfort, not just speed, not just personalization, but agency.

Do I still know what I believe? Can I still make effortful choices? Can I still tolerate friction? Can I still form my own interpretation? Can I still build real relationships? Can I still choose when not to use AI?

These may become some of the most important wellbeing questions of the next decade.

AI should be a partner, not a replacement

Llewellyn offered a helpful continuum: on one end, AI supports us. On the other end, AI substitutes for us.

Support looks like this:

• AI helps me reflect, but I still decide.
• AI helps me brainstorm, but I still think.
• AI helps me prepare for a hard conversation, but I still have it.
• AI helps me understand my stress patterns, but I still take responsibility for changing them.
• AI helps me learn, but I still wrestle with the material.

Substitution looks different.

• AI tells me what to think.
• AI decides what matters.
• AI gives me emotional relief without growth.
• AI replaces difficult conversations.
• AI removes all friction.
• AI becomes the place I go instead of developing the capacity I need.

That is not flourishing. That is dependency dressed up as support and this is where leaders have a responsibility.

If your company is implementing AI, the question is not only:

“How much productivity can we unlock?”

It is also:

“What human capacities might this system erode?”

That question changes the conversation because every technology makes something easier,  but it may also make something harder.

It may make it easier to write, but harder to find your voice.

Easier to generate ideas, but harder to think originally.

Easier to monitor work, but harder to build trust.

Easier to access support, but harder to ask another human being for help.

The wise leader asks both questions.

What does this make easier? What does this make harder?

The future of wellbeing must be psychologically designed

One of the most important insights from my conversation with Llewellyn was this: many AI wellbeing tools are being built by people with good intentions, but without a deep psychological architecture underneath them.

• They may have beautiful interfaces.
• They may have strong engineering.
• They may even have compelling content.

But if there is no theory of change, no behavioral model, no ethical guardrails, and no understanding of what actually creates human flourishing, we should be cautious. Llewellyn described seeing systems built by technically excellent teams, but around concepts they did not fully understand such as loneliness, wellbeing, and behavior change.

This is not a criticism of technologists, it is an invitation to partnership.

Psychologists, coaches, organizational practitioners, behavioral scientists, ethicists, and technologists need to build together.

Not as advisors added at the end, as co-architects from the beginning.

Because wellbeing is not just a content library. It is not just a chatbot. It is not just personalization.

It is the living interaction between a human being, their environment, their relationships, their choices, their history, their nervous system, their work, and their sense of meaning.

If we reduce wellbeing to prompts and nudges, we will miss the human and if we miss the human, we may scale harm faster than we scale healing.

A new model for AI-era flourishing.

Llewellyn introduced a powerful frame for psychologically safe AI: awareness, interpretation, intention, action, relational agency, and autonomy. At its heart is a profound idea: the future of wellbeing is about preserving our ability to write our own life stories.

That line stayed with me because the deepest promise of wellbeing is not just feeling better.

It is becoming more fully alive. It is having the awareness to see what is happening.

The interpretation to make meaning for yourself. The intention to choose what matters. The action to live those values.

The relational agency to stay connected to real people.

And the autonomy to decide when technology serves you and when it does not.

This is beautifully aligned with how we think about flourishing at Happiness Squad through our PEARL model.

PEARL was inspired by the way pearls are formed in oysters. An irritant enters. The oyster does not remove it. It transforms it.

That is work, every day brings irritants – pressure, conflict, ambiguity, change, uncertainty.

The question is whether those irritants become burnout or become the raw material for growth.

In the age of AI, PEARL becomes even more important.

Purpose: Use AI to reconnect people to meaningful contribution, not just speed. Before deploying AI, ask: How will this help people do work that matters?

Energy: Use AI to remove low-value friction, not increase always-on intensity. If AI simply helps us do more work faster without recovery, we will accelerate burnout.

Awareness: Teach people to notice how AI is shaping their thinking, attention, emotions, and choices. Awareness is the first defense against unconscious dependency.

Relationships: Protect real human connection. Use AI to prepare for conversations, not avoid them. Use it to strengthen trust, not replace care.

Leadership: Equip managers to become wise stewards of AI adoption. The manager’s role is not just to drive usage. It is to protect agency, dignity, learning, and human growth.

Three practices to start now

Here are three simple practices I would recommend for any leader, team, or individual navigating wellbeing in the era of AI.

1. First, audit your AI tools.

Write down the three or four AI tools you use most. What do they do? How do they make recommendations? What are they optimized for? What biases or limitations might they have? Llewellyn suggested comparing your assumptions with available model cards or documentation, not because everyone needs to become a technical expert, but because awareness matters.

2. Second, practice deliberate friction.

Once a week, do something without AI that you would normally outsource. Write the first draft yourself. Navigate without GPS. Think through a strategy before asking the tool. Struggle a little. Friction is not always the enemy. It is often how capability forms.

3. Third, treat AI outputs as hypotheses, not instructions.

When AI gives you advice, do not simply accept it. Ask: Does this fit my values? What perspective is missing? What would someone who disagrees say? What do I know from my lived experience that the system does not?

That is how we stay in relationship with the tool without surrendering ourselves to it.

The choice ahead

AI can become one of the greatest accelerators of human flourishing we have ever seen, or it can become another system that captures attention, increases dependency, and quietly erodes the capacities we need most.

The difference will not be the technology alone. It will be the intention behind it.

The design choices inside it. The leadership around it. The human wisdom we bring to it.

The future of wellbeing in the era of AI is not about choosing between humans and machines.

It is about making sure machines help humans become more human. More aware, more purposeful, more energized, more connected, more capable.

That is the work ahead and it may be one of the most important leadership responsibilities of our time.

Listen to the full conversation on the Flourishing Edge podcast.

Learn more about Llewellyn Van Zyl. on LinkedIn.

Listen to the podcast with Ashish and Llewellyn Van Zyl. above, You can also listen on Apple Podcasts.

Access and subscribe to all of the episodes of the Flourishing Edge Podcast here.

Make Flourishing Your Competitive Edge.