Dan Garrison leads Slalom’s overall AI strategy and vision with a focus on bringing together the company’s talented capabilities, markets, and technology partnerships teams to help customers move from AI ambition to execution. An experienced technologist and pragmatic business leader with more than 30 years of industry experience, Dan has led complex transformations, built and scaled high-performing global organizations, and advanced emerging technologies including AI.
Recently, in an exclusive interview with CIO Magazine, Dan shared insights into how his journey began with an unglamorous but inventive textbook delivery startup held together with tape and an early consulting project organizing fragmented customer data for a large shipping company, two experiences that revealed how digital could reinvent customer experience. On AI trends, Dan noted that the biggest shift in the next five years will be in people, not technology, as roles dissolve, AI becomes ubiquitous plumbing under every workflow, and judgment becomes the scarcest skill. Looking ahead, he urges future leaders to be human first and deliberately protect the struggles that build judgment instead of letting AI become a dependency. The following excerpts are taken from the interview.
Hi Dan. What was the first experience in your career that made you realize technology could truly transform how businesses succeed?
I can think of two moments, and the first one is not glamorous. Before I got into consulting, I started a company selling textbooks online. A student could go to our site, select the classes they had registered for, and see exactly which books that professor had assigned for each one. The orders printed overnight on a dot matrix printer in the back of the campus bookstore. In the morning, an employee tore the sheet off the printer, used it as a pick list, pulled the books off the shelf, loaded them into a milk crate bolted to the front of a bike, and rode them out to the dorms and apartments. Payment was a manual card swipe at the door. It was held together with tape, and it worked. That system we created was the first time I watched a website innovate on a customer experience that had belonged to brick-and-mortar retail for a century.
The second moment was my first project as a consultant, at a large shipping company. I worked on an enormous program with close to a thousand people, and it was my team’s responsibility to organize everything the company knew about a customer into a single place. While that’s standard practice now, companies in the 1990s weren’t built to operate with that level of customer detail. Customer data lived in fragments across a dozen systems and nobody had ever seen it assembled. Once my team was able to get it all in one place, it became the core that every other engine in the company’s program was built around. Same lessons in the promise of digital innovation, but with two different scales that highlighted the potential for the future.
You’re known for turning ambitious digital ideas into real results. What part of that process do you personally find most energizing?
Assembling a group of experts and framing the problem in a way that brings out the strongest thing each of them has. Cognitive empathy is the skill that makes that possible. You must understand how someone else is actually thinking about the problem, rather than thinking about how you would approach it.
Years ago, my team was a year into designing a large multinational digital program, and heading into a launch window that would not come around again for months if we missed it. The cutover sequence had to begin weeks before the launch itself. To get ahead of the risk, we ran an open scenario session, where every level was invited, and no ideas were bad – we just wanted to discuss what everyone was seeing and what could go wrong. We covered a wall in sticky notes, and I sorted most of the topics onto the side that did not need further discussion.
One of those notes came from a very quiet, very sharp junior technologist who had raised a hand almost apologetically. They had spotted a pattern in the data movement that the vendor could not explain, and if it hit while we were syncing terabytes of data, the customer experience would break. I had already moved that note to the discard side. Something about the observation pulled me back. I sat down with the technologist and they walked me through what they were seeing, and they already had the mitigation worked out. Nobody had asked them for one.
The scenario happened. We had the fallback in place and stayed on schedule. The customer’s CIO told me afterward it was the cleanest cutover of a program of any size they had been through. Running the session was the easy part. What mattered was going back to a note my own triage had thrown away.
AI is moving from pilots to production at scale. What’s the single biggest shift you expect to see in how enterprises actually use AI in the next 5 years?
I believe the biggest shift will be in people, not in the technology.
AI is already redefining roles that companies have had to do the same things the same way for decades. The line between a technologist, a strategist, and a creative is dissolving, and it is not coming back. What I expect to see is that a person’s core capabilities rise to the surface and get supplemented by the technology, which puts people into roles they never thought they would be eligible to play in, let alone succeed at. Someone whose real strength is judgment about customers has spent twenty years constrained by whether they could build the model. That constraint is going away due to AI.
There will be plenty of quieter shifts underneath. Truly innovative technology becomes ubiquitous and ends up underneath every significant business process, every workflow, every risk assessment. Those are the ones the analysts will write about. But they are plumbing. The visible change, the one leaders will actually have to manage, is the relationship between their people and the technology, and how quickly the definition of a job moves out from under someone.
The talent landscape is shifting dramatically. What skill set will be most in-demand for technology professionals in this AI-driven decade?
Judgment. Specifically, knowing when the output in front of you is wrong.
The easy answer would be to say ‘curiosity and continuous learning’, and it is true as far as it goes. The tooling looks markedly different from what it did twelve months ago, much less thirty-six, and that curve is not flattening. Anyone who is not comfortable learning and relearning at that pace is going to struggle.
But learning fast is table stakes now. The scarcer thing is the ability to look at a confident, well-formatted answer and have the intuition to spot what it’s missing. That skill is not abstract reasoning. It comes from having done the underlying work yourself, often badly, long enough to develop instincts about where things break. I can tell when a data model is going to cause problems in production because I have personally shipped ones that did.
That is what I would tell technology professionals to protect. Get fluent with the tools, absolutely. But do enough of the work by hand that you have something to check the machine against. The people who will be most in demand are the ones who can move at the speed the technology allows while still being able to say, with reasons, that this particular answer is wrong.
Leaders are shaped by what they read, watch, and learn from. What book, podcast, or talk has influenced you the most in the last few years?
Nexus, by Yuval Noah Harari. It reframed how I think about the actual risk in front of us with AI. His argument is that the story of human power is really the story of information networks, and that the question worth asking about any new one is not how intelligent it is but what it does to our ability to self-correct. That has become the lens I bring to almost every AI conversation I have with a customer.
I read Sapiens before it and The Coming Wave alongside it, and the three of them fit together into something coherent about where we are.
Outside of business, I have a long-running passion for theoretical and particle physics, which is mostly unrelated to my day job and probably the reason I enjoy it. Sean Carroll’s Something Deeply Hidden is the one I keep going back to. There is something clarifying about spending time with a field where the honest answer to a basic question is still that we do not know.
My reading habit itself has changed more than the reading list has. I used to rely on news aggregators. Now I follow specific topics, AI, bioengineering, quantum computing, and set aside time daily to scan headlines and queue up what is worth reading properly.
If you could have dinner with any innovator, past or present, who would it be and what’s the first question you’d ask them?
Richard Feynman. Whether he counts as an innovator is arguable, but he had something rarer than invention: a refusal to fool himself. He wrote that the first principle is not to fool yourself, and that you are the easiest person to fool. He built his whole method around that.
The first question I would ask him is: “How can you tell the difference between understanding something and having learned to describe it convincingly?” To me, that demonstrates true mastery of a topic.
That sounds like a physics question. It is the central management question of this decade. We now have tools that produce fluent, confident, well-structured explanations of anything, and an entire generation is about to build its professional instincts alongside them. Feynman spent his career on exactly that distinction, and he had tests for it. He would rebuild a result from scratch to see whether he actually had it. He famously judged an explanation by whether he could give it to a freshman without the jargon holding it up. My personal speaking style aligns with this today. I can go deep on a number of technology topics, but if you spend time with me, you will eventually hear me say something along the lines of, “Let’s talk about this at a ducks-and-bunnies level of detail.” It never ceases to amaze me how some people who are experts in their craft are unable to talk about it with anyone who is not on the exact same level they are.
Back to Feynman, I would want to know what his tests would be now, when the explanation arrives already polished and there is nothing visibly missing. That is the problem I do not have a good answer to, and I think he would have had one.
Everyone has a go-to source of energy outside of work. What hobby or activity helps you reset when things get intense?
Brazilian jiu-jitsu. I get to the gym far less than I would like, once or twice a week when the calendar allows, but it does something nothing else does. It is a completely primal sport, and for the length of a round there is no possibility of thinking about anything else. Someone is trying to submit you. That focus is not available to me anywhere else, and it counteracts a life spent in front of a laptop and in row 23 of an airplane.
The other one is mindfulness. I have practiced meditation for ten or twelve years now, and I try to meditate every day. What is instructive is what happens when I do not. When work gets busy enough that the practice falls off the calendar, it catches up with me every time. I will not claim it puts me at my best. But it clearly makes me better.
Both of them work for the same reason. They require full attention on something that is not a strategy or technical problem, which is the only real way I have found to put one down.
What is your biggest goal? Where do you see yourself in 5 years from now?
Honestly, my biggest goal is raising two great kids. My real ambitions sit closer to family, peace of mind, and who I am as a person versus anything on a career track. I recognize that is not the expected answer from a Chief AI Officer, and it’s the true one.
On the professional half of the question, we are moving into a chapter where technology genuinely changes the nature of what a business is, not just how efficiently it runs. I expect to be at Slalom in five years, working on that. What I do not expect is for this role to look anything like it does today. Chief AI officer is a title that barely existed a few years ago, and there is no reason to assume the shape of it is settled. The parts that feel central right now (standing up capability, building governance, getting the organization fluent) may be finished work by then, and something I cannot currently name will have taken their place.
That is the part I am looking forward to. I would rather be in a role that keeps changing under me than one I have already figured out.
Looking ahead, what’s the one piece of advice you would give to future leaders who will grow up leading in an AI-first world?
Be human first. At Slalom, we talk about this being “fiercely human”.
I understand what AI-first is meant to convey, and I do not think it is the right instruction for a leader. The technology should be an extension of human capability. The risk we are actually running is that it becomes a dependency instead, and that shift happens quietly enough that nobody recognizes it until it is painfully too late.
The specific version of this that concerns me is what I think of as a skipped generation. People entering the workforce now will never know a professional world without these tools. If organizations do not deliberately structure roles, learning paths, and challenges around that fact, we will produce a cohort that is extraordinarily productive (due to AI) but has no foundation underneath it. I built my instincts the hard way, writing the email myself, rebuilding the deck myself, sitting through the difficult conversation with no way around it. That was inefficient. It was also where the judgment came from. It made me who I am today.
I am not arguing for making people do work a machine can do better. I am arguing that leaders have to decide, on purpose, which struggles are still worth keeping, because they are the ones that build the person. Nobody will do that by accident, and the efficiency case will always argue against it. I am not alone with this concern or this thinking, I am sure.
That is the job for the next generation of leaders. Not adopting AI, which will happen regardless. Protecting the conditions that produce people worth putting in charge of it.