Challenge Accepted series

September 2nd, 2026 | Hosted by  Klil Nevo, Juno Journey
Expert: Dr. Cara Antoine

---

ICYMI: The Human Transformation in the Intelligent Era: What Will the Future Require of Us to Become?

AI is evolving incredibly fast.

But perhaps the more important transformation isn’t happening inside the technology.

It’s happening to us.

That was the central idea behind our latest Challenge Accepted session with Dr. Cara Antoine, Executive Vice President and Chief Technology, AI & Innovation Officer, who has spent more than 30 years leading technology, transformation, and innovation across complex global organizations.

Rather than asking what AI will be capable of next, Cara challenged us with a different question:

How do we need to evolve to become what the future requires of us?

And for HR, L&D, Talent, and business leaders, that question may matter far more than learning the latest AI tool.

Technology has always augmented humans. What’s different now is the speed.

Cara started by taking us much further back than ChatGPT.

Humans have always built tools to extend what we’re capable of doing.

Stone tools helped us survive. The wheel helped us move farther and faster. The printing press transformed how knowledge traveled. Electricity, the telephone, and computing changed how people lived and worked.

The pattern itself isn’t new.

What has changed dramatically is the speed of adoption.

Technologies that once took decades to spread can now reach enormous numbers of people almost immediately. AI is already embedded in the tools people use every day, and organizations are racing to understand where it belongs.

As Cara put it:

“The speed of possibility has outpaced the pace of preparation.”

And that creates a very different challenge for organizations.

Keeping up can no longer mean periodically training people on a new system.

The technology will continue changing.

People need to develop the ability to continuously change with it.

FOMO is not an AI strategy

More than 70% of organizations, Cara noted, are exploring what AI could mean for their businesses.

But experimentation and transformation are very different things.

The initial excitement around generative AI pushed organizations to launch pilots, proofs of concept, and prototypes everywhere. Yet only a small fraction of those experiments have made it into production at meaningful scale.

Part of the problem is FOMO: fear of missing out.

Organizations see competitors investing in AI and conclude that they need AI everywhere too.

Cara’s advice was much more deliberate:

Don’t start with AI. Start with the problem.

There are processes where AI can create enormous value through prediction, automation, generation, and orchestration.

There are others where simpler technology — or simply a better process — may be the better answer.

The goal isn’t maximum AI.

It’s understanding what tool belongs in what part of the work.

Mindset → Skill set → Toolset

One of the most practical frameworks from the session was the order in which AI transformation should happen:

Mindset → Skill set → Toolset.

Not the other way around.

1. Mindset

Our relationship with work is changing.

Cara described a shift from humans doing roughly 80% of the work and directing systems for the remaining 20%, toward a world where people may increasingly spend more time directing, evaluating, and deciding, while technology performs more of the execution.

That requires people to rethink what their value actually is.

2. Skill set

AI capability can’t sit inside one specialist team.

Every function and every role will increasingly need to understand how AI changes the way their work gets done.

The question for organizations therefore becomes less about whether employees have “AI skills” in general and more about whether they understand where and how to work with AI inside their actual roles.

3. Toolset

Only then should organizations decide what technology belongs inside the process.

That distinction matters because transformation doesn’t come from handing everyone another AI tool.

It comes from redesigning the work.

Stop asking which jobs AI will replace. Start breaking down the work.

One of Cara’s strongest examples involved taking a large business process and breaking it down into its smallest components.

A process like order-to-cash can contain hundreds of individual steps.

Once those steps become visible, leaders can ask:

Where do we need a human?

Where would traditional automation work?

Where would generative AI help?

Where could an AI agent take responsibility?

And where is human judgment essential?

The answer usually isn’t that AI takes the entire process.

Instead, the work begins to look like a wave: humans, automation, generative AI and agents moving in and out at different points depending on what each part of the process requires.

That gives leaders a much more useful question than:

“Can AI do this job?”

Ask instead:

“Which parts of this work should be done by humans, which should be augmented, and which should be automated?”

Human in the Loop isn’t enough

Cara offered three ways to rethink one of the most familiar phrases in AI: Human in the Loop.

The first is the one we already know:

Human in the Loop.

People remain involved in AI-enabled processes.

But Cara pushed the idea further:

Human in the Lead.

Humans shouldn’t simply be checkpoints inside an automated process. They remain accountable for where the process is going, why it exists and what decisions ultimately get made.

And then came perhaps the most memorable version:

Human with the Loupe.

A loupe is a magnifying glass.

The role of the human is to inspect.

To question.

To look closely at what the system produces.

To understand the context the AI doesn’t.

To apply judgment.

This becomes especially important because AI operates probabilistically. It can generate something plausible without it necessarily being true.

The more AI contributes to our work, the more valuable human judgment becomes.

Data is history in numbers

That responsibility becomes even more important when we consider what AI learns from.

Cara made a simple but powerful observation:

“Data is basically history in numbers.”

If history is incomplete, the data will be incomplete.

If certain experiences were never recorded, they won’t appear in the dataset.

If the people creating algorithms lack diversity of background, perspective or experience, those limitations can become embedded into the systems we build.

This makes critical thinking, judgment and diversity much more than “human skills” sitting beside an AI strategy.

They become part of the AI strategy itself.

Humans need to be able to challenge what AI produces rather than simply accept it because the output looks confident.

The future organization will include humans and agents

Today, many organizations are experimenting with agents inside individual workflows.

But Cara expects this to evolve significantly.

Agents may move from supporting one process to collaborating across multiple processes, then orchestrating work across an enterprise, and eventually interacting with agents belonging to customers, suppliers, and partners across entire ecosystems.

That means organizations will increasingly need to answer questions they haven’t had to answer before.

How do humans and agents divide responsibility?

Where does accountability sit?

How much autonomy should an agent have?

How do people learn to collaborate with something that may increasingly look like another member of the team?

And what happens to the traditional organizational pyramid?

Cara was clear on one particularly important point: organizations should not assume that AI makes junior talent disposable.

Entry-level employees are how organizations develop future generations of talent.

The work they do may change dramatically.

Their relationship with AI may be completely different from the generations before them.

But eliminating the bottom of the talent pipeline could also eliminate the people organizations need to lead them in the future.

Before you automate, use ESSA

During the Q&A, Cara shared another framework worth keeping:

Eliminate → Simplify → Standardize → Automate.

In that order.

Before adding AI to a process, ask:

Does this step need to exist at all?

Can the process be simplified?

Can we standardize it?

And only then: should we automate it?

It’s a useful reminder at a moment when “add AI” is becoming the default answer to almost every organizational challenge.

Automating a broken process doesn’t necessarily make it better.

Sometimes it simply makes the broken process move faster.

The real AI transformation is human

Perhaps the biggest takeaway from the session was that none of us can treat AI transformation as something happening somewhere else in the organization.

It isn’t an IT initiative.

It isn’t something the AI team handles.

And it isn’t simply about teaching employees how to prompt ChatGPT.

AI will change how decisions are made, how processes operate, how teams are structured and what people spend their time doing.

Which means every organization has another transformation to manage alongside the technological one:

the transformation of its people.

The skills we value will change.  The work we perform will change.  Our relationship with technology will change.  And our own roles will change with them.

Cara ended with a challenge worth carrying into every AI conversation:

Instead of obsessing over what the technology will become next, ask:  What will this require us to become?

Because the organizations that thrive in the Intelligent Era may not simply be the ones with the best AI.

They’ll be the ones whose people learned how to evolve with it.