Thought Leader Q&A: Dimitris Tolis

Transferring From Static eLearning And AI-Generated Content material To Competency-Pushed Studying Experiences

With greater than 25 years of expertise in Studying and Growth, Dimitris Tolis is the Founder and CEO of Human Asset, the place he has led the design of customized eLearning, studying academies, and AI-powered studying options for European businesses reminiscent of EUAA, CEPOL, EUDA, and worldwide organizations, such because the Council of Europe, ESM, United Nations ITU. As a Senior Tutorial Designer, Licensed Govt Coach, and AI Researcher on the College of Turku Finland, he brings collectively Tutorial Design, neuroscience, and academic know-how to create studying experiences which might be extra human-centered, adaptive, and practice-based. By means of initiatives reminiscent of gAImify Hub, he’s serving to shift the dialog from sooner content material manufacturing to extra significant studying design. At the moment, he speaks with us in regards to the alternatives, dangers, and way forward for AI in office studying.

Based mostly in your expertise, what are the dangers of present AI use in studying, and the way can they hinder significant L&D journeys?

One of many greatest dangers is that AI is fixing the improper drawback in studying. It helps us create content material sooner, however pace alone doesn’t enhance studying. As an alternative, it could possibly result in content material mediocrity at scale: extra slides, quizzes, and modules, however with weaker educational depth, much less originality, and a poorer learner expertise. It might probably additionally create what I name a “little God” impact: the phantasm that as a result of content material could be generated immediately, significant studying has additionally been designed. With out robust Tutorial Design, this rapidly results in content material inflation and decrease high quality.

A second danger is cognitive offloading mixed with overdependence on AI. When learners obtain on the spot solutions, simplified summaries, and predictable suggestions, they might have interaction much less deeply. Vital pondering, reflection, and judgment can weaken over time, as we already discover taking place.

One other critical danger is AI hallucination. Massive language fashions can produce outputs that sound fluent, assured, and credible, even when they’re inaccurate, deceptive, or utterly false. In a studying context, that’s particularly harmful, as a result of learners might belief the reply just because it’s effectively written. If that is mixed with weak overview processes, poor prompts, or no educational guardrails, AI can unfold confusion relatively than assist understanding.

So significant L&D journeys could be hindered when AI makes studying sooner but in addition flatter.

My view is optimistic, although: these should not causes to step again from AI. They’re causes to design it higher.

What are a few of the most ignored alternatives for AI in studying, and why ought to organizations shift from content material era to significant studying expertise design when implementing this rising know-how?

One of the crucial ignored alternatives is that AI might help us transfer from info supply to functionality constructing. Most organisations nonetheless use AI primarily to generate content material sooner. Nonetheless, the true worth lies in designing studying experiences which might be extra adaptive, extra contextual, and extra practice-based.

A great instance is the position of adaptive quizzes. Too typically, quizzes merely test recall. With AI, they will turn out to be a part of the educational course of itself. The extent of problem can shift dynamically, weaker areas could be strengthened, and customized suggestions can information the learner ahead. That makes quiz apply extra developmental and far nearer to actual studying.

One other main alternative is open-ended apply with personalised suggestions. Many essential office expertise, reminiscent of interviewing, giving suggestions, teaching, dealing with battle, and so forth., can’t be developed by means of multiple-choice questions alone. Learners want to reply in their very own phrases, make judgements, and mirror on their decisions. AI can assist this by means of AI teaching personas that present extra focused suggestions on readability, reasoning, empathy, tone, and intent.

This issues as a result of significant studying is just not created by making issues simpler. It’s created by providing the appropriate problem with the appropriate assist. Aristotle’s perception nonetheless holds true: studying requires effort. Actual studying and growth occur when learners are challenged. And Bloom’s 2 Sigma analysis reminds us of the worth of personalised steerage. AI offers us an opportunity to convey each collectively at scale for the primary time in human historical past.

Lastly, AI creates an essential alternative for customisation. As an alternative of one-size-fits-all coaching, studying could be formed across the organisation, the position, the competencies, and the context. That’s the reason organisations ought to shift from content material era to significant studying expertise design.

What’s the significance of human-centered AI and human-in-the-loop approaches when constructing competency-driven studying experiences?

Hallucinations, the black-box nature of LLMs, and what I typically name the “immediate and pray” strategy are precisely what make AI dangerous in studying. If we merely ask a mannequin to generate content material, suggestions, or evaluation with out robust construction, we might get outputs that sound fluent and convincing, however should not essentially correct, related, or pedagogically sound.

That’s the reason human-centred AI and human-in-the-loop are so essential, particularly in competency-driven studying. They assist transfer AI from improvisation to disciplined design.

With the appropriate structure, we will maintain AI targeted by means of particular competency frameworks, grading rubrics, clear educational targets, guardrails, and moderation logic, and naturally, human overview and approval. This makes a significant distinction. As an alternative of letting AI wander, we information it towards what issues: the talents, behaviours, and requirements we truly need learners to develop.

In sensible phrases, which means AI can assist the expertise by producing apply, suggestions, and adaptation, whereas people stay answerable for high quality, alignment, and belief. The result’s a studying setting that’s extra dependable, extra clear, and extra developmentally significant.

For me, that is the true worth of a human-centred strategy: it makes AI extra reliable, but in addition extra helpful. It permits us to learn from pace, responsiveness, and personalisation with out dropping pedagogical integrity. In competency-driven studying, that steadiness is crucial.

Are you able to describe a consultant AI-powered studying transformation use case out of your work?

Sure. A consultant instance from our work includes a significant regulation enforcement academy in Europe, the place we’re co-designing an AI-powered Practice-the-Trainers capability constructing program targeted on serving to trainers strengthen their educational design and supply expertise.

What makes this case particularly significant is that the course is designed round a twin function: to cut back AI dangers, reminiscent of hallucinations, overreliance, weak judgment, and poor educational use—and on the similar time to unlock AI alternatives in additional personalised, adaptive, and practice-based studying.

The transformation is just not about including AI on prime of a standard course. It’s about redesigning the educational expertise itself. We’re utilizing AI-assisted course design with structured templates, customisation to the academy’s context and coach roles, adaptive quizzes that assist apply relatively than easy recall, open-ended eventualities with coaching-style suggestions, and AI avatar simulations that permit trainers to rehearse practical conversations and facilitation moments. We additionally use competency frameworks, rubrics, and human-in-the-loop overview to maintain the expertise reliable and aligned with the academy’s requirements.

What I discover most fun is that this type of venture strikes AI from content material era to functionality constructing. For me, that could be a very robust instance of AI-powered studying transformation: not sooner content material, however higher studying design.

Is there a current growth venture, product launch, or one other initiative you’d prefer to share with our readers?

Sure, I’d be very glad to share gAImify Hub, one among our most essential current initiatives at Human Asset.

gAImify Hub is our AI-powered, gamified studying platform designed to assist organisations create studying that’s extra adaptive, extra practice-based, and extra carefully linked to actual office efficiency. What makes it particularly essential to us is that it displays a really deliberate philosophy: AI shouldn’t merely assist us produce content material sooner. It ought to assist us design higher studying experiences.

The platform brings collectively AI-assisted course design, contextual customisation across the organisation and the position, adaptive quizzes, open-ended eventualities with coaching-style suggestions, real-time AI avatar simulations, and gamified studying journeys. So as an alternative of counting on static eLearning alone, organisations can create experiences the place learners suppose, reply, practise, mirror, and enhance.

A key a part of the innovation can also be the human-in-the-loop strategy. AI helps the design and the learner expertise, however studying professionals stay answerable for overview, refinement, and approval. For us, that’s important. It retains the expertise extra reliable, extra related, and extra aligned with actual studying targets.

Simply as importantly, gAImify Hub has been designed with a robust emphasis on moral AI and compliance. That features accountable use of AI, clear human oversight, and a focus to necessities round knowledge safety, belief, and governance, together with GDPR and broader Authorized readiness. We see this as a needed basis for innovation in studying, not as an afterthought.

These improvements could be utilized in two methods: construct new adaptive studying experiences with gAImify Hub or improve present SCORM programs with inSCORM AI.

What do you suppose the long run holds for AI in adaptive studying academies?

I consider the way forward for AI in adaptive studying academies is extraordinarily promising, however it’s going to rely on the alternatives we make now. The way forward for AI in training won’t be determined by who produces probably the most content material, however by who designs probably the most significant studying.

The strongest academies will use AI to maneuver past static programs and create studying ecosystems which might be extra adaptive, extra practice-based, and extra linked to actual functionality growth. They won’t merely ship info. They’ll assist learners suppose, practise, mirror, obtain suggestions, and enhance over time.

For me, one precept is crucial: AI ought to make studying tougher and interesting, not simpler within the improper manner. It shouldn’t scale back effort or encourage passive dependence. It ought to assist create the correct of problem, with the appropriate assist, on the proper second. That’s the place adaptive studying turns into actually highly effective.

I additionally consider academies will turn out to be way more clever in how they reply to learners. We’ll see stronger use of adaptive evaluation, open-ended eventualities, simulation-based apply, and suggestions loops that make growth extra seen and extra personalised.

On the similar time, one of the best academies will stay deeply human-centred. They’ll mix AI with robust pedagogical design, moral guardrails, and human judgment.

So, I’m optimistic. I feel AI offers academies an actual alternative to evolve from content material libraries into dwelling environments for development, reflection, and efficiency. That, to me, is the extra inspiring future.

Wrapping Up

Thanks a lot to Dimitris Tolis for sharing his insights on the potential dangers and alternatives of utilizing AI to create personalised, adaptive studying experiences. If you would like to delve deeper into this matter, take a look at Human Asset’s information, AI in Workplace Learning: From Content Generation to Meaningful Learning Design.