Last week at the Association for Talent Development (ATD) Conference in Los Angeles, I had the opportunity to speak with senior Learning & Development leaders from across industries. From fast-growing SMBs to Fortune 500 enterprises the conversations surfaced one question again and again:
“How do we make human power skills visible in a way that is measurable, developable, and tied to real business outcomes?”
The conversations were remarkably consistent. L&D leaders and CHROs are under increasing pressure to demonstrate measurable ROI from leadership development, training, and workforce transformation initiatives. Yet many of the capabilities organizations care about most, creativity, judgment, communication, adaptability, leadership, and collaboration, remain difficult to measure with traditional approaches.
At the same time, these capabilities have never been more important.
As generative AI rapidly reshapes work, organizations are discovering something profound: technical AI fluency alone is not enough. The future belongs to people who can combine AI with human judgment, creativity, ethical reasoning, adaptability, and collaborative leadership.
And yet this combination of skills is precisely what traditional talent systems struggle to see.
While in Los Angeles, I made my annual visit to the Griffith Observatory. The Observatory never fails to inspire reflection on science, discovery, and the power of making the invisible observable.
Earlier in my career, as a PhD student, I worked at another observatory where I had the privilege of helping students, families and visitors experience that same sense of wonder: transforming distant, abstract phenomena into something directly observable and deeply human. Standing again beneath the Griffith Observatory dome, I was reminded how profoundly visibility changes understanding and how similar that mission feels to the work we are doing today at Ignis AI.
Standing there overlooking the city, watching visitors line up to peer through the Observatory’s historic Zeiss refracting telescope, I kept thinking about the Observatory’s mission: transforming the “visitor into observer.”
That phrase stayed with me.
For decades, organizations have treated human capabilities largely as inferred traits rather than directly observable behaviors. We infer leadership from job titles. We infer communication from résumés. We infer creativity from educational pedigree or reputation.
But inference is not measurement.
The Griffith Observatory changed humanity’s relationship with the sky by making observation accessible at scale. More people have looked through its telescope than any other telescope on Earth. Suddenly, what once felt distant and abstract became visible and shared.
I believe we are at a similar moment in workforce transformation.
Essential human capabilities such as creativity, leadership, communication, and collaboration are no longer hidden traits.
They are observable behaviors that can be measured and developed.
That realization sits at the core of what we are building at Ignis AI.
At Ignis AI, we call this approach Power Skills Practice and Proof.
The idea is simple but transformative:
Organizations should not have to guess whether employees are developing the capabilities required for the future of work. They should be able to observe growth, measure progress, and connect development efforts to meaningful outcomes.
Historically, that has been difficult because power skills are fundamentally different from technical skills.
You can test whether someone knows a software platform or a compliance process with a traditional assessment. But how do you measure creativity under ambiguity? Judgment under pressure? Adaptability when circumstances change? Leadership when there is no obvious right answer?
The challenge becomes even more complex in the age of AI.
Today, AI can help generate polished outputs, presentations, strategies, analyses, and recommendations. Surface-level quality is no longer enough to tell us where the human contribution ends and the AI contribution begins.
That changes everything about assessment and talent development.
The most important question is no longer:
“Can someone produce a good answer?”
It is:
“How does someone think, adapt, collaborate, evaluate tradeoffs, and generate differentiated judgment in partnership with AI?”
That is where the most interesting work lies ahead.
One of the themes I heard repeatedly at ATD was frustration with talent systems built primarily on inference.
Resumes, job histories, credentials, engagement metrics, and AI-generated skill profiles can provide useful signals. But they often tell us more about exposure than actual capability.
Two people can have identical résumés and dramatically different levels of creative thinking, leadership effectiveness, or adaptability.
This distinction between inferred skills and measured skills is becoming increasingly important in AI-enabled work environments - and it is one worth understanding in depth. It also explains why the data feeding your talent systems matters as much as the systems themselves.
Organizations need ways to observe how people actually reason, communicate, and solve problems in realistic contexts.
That is why we believe the future of assessment and development is performance-based, contextual, and embedded in authentic work activity.
Not static tests.
Not one-time snapshots.
Not personality proxies.
But continuous signals of capability emerging through real-world practice.
Our recent research at Ignis AI explores how advances in agentic AI, computational psychometrics, and scenario-based assessment can make these capabilities visible at scale.
In our work, creativity is not treated as an abstract personality trait. It is operationalized as observable behavior: the ability to generate non-obvious alternatives, reframe problems, adapt under changing constraints, and sustain differentiated thinking in complex environments. Our research on measuring creativity in the age of generative AI - presented at BIGAI@MIT - explores exactly how to distinguish genuine human originality from AI-assisted output.
This matters deeply for workforce transformation.
As AI systems become increasingly capable, organizations face a growing risk of homogenization. AI can improve fluency and productivity, but it can also push individuals and teams toward increasingly similar outputs and patterns of thinking.
The differentiator will not simply be who uses AI.
It will be who can think differently with AI.
Who can challenge assumptions.
Who can synthesize across domains.
Who can adapt under ambiguity.
Who can bring human creativity and judgment into partnership with intelligent systems.
What excites me most is that this moment requires more than simply building new assessments or adding another leadership training program.
It requires rethinking the system itself.
The future of talent development will not be built around episodic training disconnected from work. It will increasingly revolve around continuous readiness, embedded development, and observable evidence of capability growth in context.
In other words:
The science behind this approach is detailed in our white paper on advances in AI and psychometrics for Power Skills measurement, and validated through our peer-reviewed study with Arizona State University and LERN, which demonstrated measurable improvements in decision-making quality and leadership effectiveness in real-world deployments.
This is just the tip of the iceberg.
The conversations at ATD reinforced that organizations across industries are beginning to recognize the same emerging reality:
The future of workforce advantage will depend not only on AI adoption, but on our ability to develop and measure the uniquely human capabilities that allow people to use AI wisely, creatively, and adaptively.
And perhaps most importantly:
To make those capabilities visible.
I’m grateful for the thoughtful conversations, questions, and challenges shared by leaders throughout the week at ATD.
The future of human capability development is becoming observable.
About the Author Dr. Yigal Rosen is Co-Founder and Chief Product Officer at Ignis AI, and an internationally recognized leader in human capability assessment at the intersection of AI, computational psychometrics, and learning sciences. Before founding Ignis AI, he held senior roles at Harvard University, BrainPOP, Pearson, and ACT, and led two landmark OECD/PISA global assessments - the 2015 Collaborative Problem Solving Assessment and the 2022 Creative Thinking Assessment - that transformed how education systems in more than 70 countries measure human capability. He holds a Ph.D. from the University of Haifa, completed postdoctoral fellowships at Tel Aviv University and Harvard, and is currently pursuing his Executive MBA at MIT Sloan School of Management.