The Promise of AI.
Why are so many organizations struggling to see measurable impact from AI?
By John Vardalos, CEO, J5 DesignEvery leader I’ve spoken with this year has felt the same push and pull: AI represents some of the most exciting possibilities we've seen in a decade, yet the path to real, sustainable results feels uncertain. For some, AI has accelerated innovation faster than they imagined. For others, it has stalled progress, created confusion, or introduced risks that are harder to manage.
No matter where an organization sits on that spectrum, one truth is becoming clear: AI is not failing us, our approach to AI is failing. And the data backs this up.
MIT Study: AI projects are falling short.
A recent MIT study revealed that 95% of generative-AI pilots fail to deliver measurable ROI. Companies are pouring resources into experimentation, yet most pilots never scale and never move the needle on impact.
The Forbes analysis of the report highlights why:
Organizations are avoiding friction, the very friction that reveals how people actually work, where pain points live, and where AI must integrate rather than simply “install.” Many teams focus on the technology itself instead of redesigning the experience around it. Too many AI initiatives start with the solution, not the problem. And when that happens, two things follow:
The technology becomes misaligned with real workflows and needs
People don’t trust, understand, or adopt what’s been introduced
This is not a technology failure. It is a design failure.
Why understanding experience matters more than ever.
To understand why adoption lags, consider a familiar example: autonomous vehicles.
A self-driving car does not need to show the passenger visual cues, the little icons of pedestrians, cyclists, nearby cars, or glowing outlines of traffic lights in order to function safely. The vehicle could navigate perfectly without ever presenting these elements on the interface.
But humans need to see them to trust the system.
Those visuals are not technical requirements. They are emotional requirements. They reassure the passenger that the vehicle sees what they see, understands the environment, and is acting with awareness. Trust is not a by-product of technology. Trust is designed. And the same principle applies to organizations adopting AI.
When teams don’t understand what AI is doing, when workflows change without being co-designed, or when users can’t connect the technology to the problems they actually face, adoption collapses. Even the most powerful AI solution cannot overcome a poor experience.
Healthcare example - AI works because the experience was respected.
Healthcare often gets used as an example of AI disruption, especially in radiology. Years ago, many believed AI would replace radiologists or MRI technologists entirely. But that never happened and the reason is revealing.
AI succeeded in MRI not because it “replaced tasks” but because it enhanced the experience for everyone involved:
Patients benefited from more consistent scans and fewer rescans
Technologists were relieved of repetitive, tedious steps and gained time for high-value clinical decision-making
Radiologists’ capacity increased, enabling departments to hire more, not fewer, experts
The human experience was not an afterthought, it was central to adoption. AI worked because the workflow, the safety needs, the emotional needs of patients, and the clinical judgement of humans were integrated into the design from the beginning. It is a model for what responsible, high-impact AI adoption can look like.
People Adopt What They Understand, Trust, and Help Create
This is the pattern across every successful AI implementation:
When users are co-creators, not bystanders
When friction is surfaced and addressed, not avoided
When the emotional experience is designed with as much care as the technical one
When the real problem is diagnosed before the solution is defined
AI becomes not just functional, it becomes adoptable. Organizations that ignore these realities may find themselves among the 95% whose pilots stall. Those that embrace them will be the ones who transform.
What This Means for J5 and the Future of Human-Centred AI
At J5, this insight guides the direction of our work, especially in healthcare where stakes are high and trust is non-negotiable. We’re now building products and services that integrate AI into the design process itself not to replace the work we do with clients, but to augment it, making consulting faster, more precise, and more impactful.
AI is helping us accelerate tasks that used to take days, so we can spend more time on what truly matters:
listening deeply, co-creating solutions with staff and stakeholders, and supporting teams through the behaviour change required for transformation.
And we’re doing it with a commitment to what has always defined J5: a belief that innovation is only meaningful when it creates kinder, more beautiful futures grounded in human experience.
The Opportunity Ahead
AI will continue to reshape industries and redefine roles. Tasks will evolve. New professions will emerge. Entire sectors will be reborn. But one thing won’t change: Transformation will succeed only when people feel seen, understood, and supported through it.
If we design for the human experience first, AI can unlock extraordinary outcomes. If we design for technology first, AI will continue to stall, no matter how powerful it becomes. The future belongs to organizations willing to slow down just enough to ask a simple question:
Are we solving the right problem? Are we solving it with the people who will live with the solution?
That is where real AI transformation begins.
