Recursive Human Improvement: What if AI helped us become the humans we've always wanted to be?

We're teaching machines to improve themselves. But who's helping humans do the same?

Something caught me out this week.

A thought had been going around in my head about recursive human improvement. I hadn't quite got around to doing anything with it.

Then, scrolling through my feed, I came across Arianna Huffington interviewing Daniela Amodei, President of Anthropic.

They were talking about AI and humanity as two partners in a dance. And the danger of one partner becoming more and more capable while the other stands still.

Then came the idea that stopped me.

Humans can recursively improve themselves, too.

Of course we can.

And I think this could be one of the most important opportunities AI presents to humanity.

We've been asking the wrong question.

Billions are being invested in making artificial intelligence progressively more capable.

Machines that learn, reason, solve increasingly complex problems and may eventually help improve their own capabilities.

Meanwhile, we're busy asking how humans can keep up.

What skills will we need? Which jobs will survive? How do we remain relevant?

Important questions.

But they're not nearly ambitious enough.

What if the greatest achievement of artificial intelligence turns out to be the recursive development of human intelligence?

Not simply helping us do more.

Helping us become more.

There's more inside you than you know.

I've believed this throughout my career.

I've seen people discover capabilities they didn't know they possessed. Leaders become people they never imagined they could be. People find the courage to do things they once thought were beyond them.

And I've always believed that my job, whether as a leader, coach or mentor, is to help people astound themselves with what they're capable of.

Because human improvement isn't simply about becoming cleverer, faster or more productive.

It isn't about endlessly optimising ourselves to keep up with machines.

It's about becoming more fully the human you've always wanted to be.

More curious. More courageous. More confident in your own judgement.

More capable of imagining possibilities, making choices and acting on them.

More authentically yourself.

And perhaps discovering an entirely new version of yourself that you didn't previously know was possible.

That's my definition of human improvement.

So what is Recursive Human Improvement?

In artificial intelligence, recursive self-improvement describes the possibility of a system using its capabilities to improve those very capabilities, potentially enabling further improvement.

But humans have always had their own version of this.

We learn something.

That changes how we think.

Our improved thinking helps us recognise something we couldn't previously see.

Which enables us to learn, develop and achieve something else.

And so the cycle continues.

Now imagine what happens when we deliberately bring artificial intelligence into that process.

Not to do our thinking for us.

But to help us think better.

To challenge our assumptions. Reveal our blind spots. Introduce perspectives we haven't considered. Help us experiment, reflect and learn.

And crucially, to develop capabilities that remain ours.

My proposed definition of Recursive Human Improvement is:

The deliberate process through which humans use experience, reflection, learning and human–AI co-creation to improve their own capacity to think, learn, judge and act, enabling each cycle of development to expand the possibilities of the next.

This is an idea I'm developing through Human Frontier Thinking® and HI:AI®.

And I believe it offers an exciting new way of thinking about human development in the age of AI.

The difference between doing better and becoming better.

Imagine a leader who uses AI to produce a brilliant strategy.

That's useful.

But imagine instead that the leader uses AI to question their thinking, explore unfamiliar possibilities, challenge their assumptions and discover a better way of making decisions.

They don't just produce a better strategy.

They become a better strategic thinker.

And next time, they can tackle an even more complex challenge.

That's the difference.

AI augmentation improves what we can accomplish with AI.

Recursive Human Improvement develops what we become capable of accomplishing next.

And I think that's where the really exciting opportunity lies.

What if we're measuring the wrong things?

We're becoming rather good at measuring AI productivity.

Hours saved. Tasks automated. Reports produced. Costs reduced.

But how often do we ask:

What has the human gained?

Are they more capable of independent thought?

Has their judgement improved?

Are they more confident navigating uncertainty?

Can they now recognise possibilities they couldn't see before?

Are they becoming more curious, creative, courageous and capable?

Or are they gradually becoming more dependent on a machine that does their thinking for them?

Because there's a very real danger that we develop increasingly intelligent machines alongside increasingly passive humans.

And that's not a future I find particularly attractive.

The human frontier isn't AI.

Over the past couple of years, I've been developing Human Frontier Thinking® the discipline of continually expanding how humans think, learn and act at the edge of change, discovering capabilities and possibilities we did not previously know existed.

At its heart is a simple belief.

Human potential is not fixed.

We can continually expand our capacity to notice, question, think independently, exercise judgement, co-create and imagine beyond our present constraints.

HI:AI® explores what becomes possible when we bring human and artificial intelligence together deliberately, with humans retaining agency, judgement and purpose.

And above both sits an ethos I call Progressus Supra Se® "progressing beyond ourselves."

Recursive Human Improvement may be the mechanism that brings these ideas together.

A continuous cycle of thinking, questioning, creating, judging, acting and learning.

Not simply repeating what we already know.

But developing the capacity to go further each time.

But there's something we mustn't lose.

Human improvement is not the same as machine improvement.

We're not trying to become faster processors or more efficient versions of ourselves.

And growth doesn't mean we must constantly strive to be someone else.

Sometimes improvement means slowing down.

Finding the confidence to trust your own judgement.

Being more present with the people you love.

Discovering what matters to you rather than what the world expects of you.

Finding the courage to be yourself.

Because the purpose of human development isn't endless optimisation.

It's greater human agency, possibility and fulfilment.

And if AI helps us achieve that, we may be looking at something far more significant than another technological revolution.

Perhaps it's time to raise our ambitions for AI.

I don't want a future in which humans simply learn to keep up with increasingly capable machines.

I want one in which the development of artificial intelligence helps accelerate the development of human intelligence, imagination and potential.

Where every generation of AI opens new opportunities for humans to discover what they're capable of.

Where we don't simply become better users of technology.

We become better thinkers, learners, creators, leaders and more fully ourselves.

Because there's more inside us than we know.

And perhaps, for the first time in human history, we have a technology that could help us discover just how much more.

We're teaching machines to improve themselves recursively.

Now let's get serious about helping humans do the same.

And here's the question I'd love us to start asking:

What if every meaningful interaction with AI left us not just with a better answer, but with a more capable human?

Carol Glover is the founder of Human Frontier Thinking® and HI:AI®, exploring how humans can expand their thinking, agency and potential in an AI-shaped world.

#HumanFrontierThinking #HIAI #RecursiveHumanImprovement #HumanPotential #FutureOfHumanity