May 2026

You’ve seen many articles that start with “AI has changed everything.” We’re not going to write that one. Not because it isn’t true; but the more interesting question isn’t what has changed. It’s whether we’ve changed with it in the right ways, for the right reasons.

At NovusNorth, we’ve been doing some deep thinking on that front. What follows is less a manifesto and more a working document: where we stand, what we’ve unlearned, and a few takes we’re willing to defend. And we reserve the right to continue to refresh this point of view!

The Toolset Has Shifted. The Work Has Not.

When generative AI began flowing into design projects and tools in earnest, there was a predictable split in reaction. One camp declared the discipline disrupted beyond recognition. Another dismissed it as a better autocomplete. Neither was right.

What we’ve actually seen is something more nuanced: the surface area of design work has expanded dramatically, while the hard part that includes making meaningful decisions on behalf of real people and solving real business problems has stayed exactly as hard as it always was. Maybe harder.

AI tools have compressed the distance between concept and artifact. Prototypes that once took days take hours. Visual directions that once required a full production cycle can be stress-tested (synthetically or live with users) in an afternoon. That’s genuinely useful. But compression isn’t the same as clarity. The bottleneck in design was never “can we make enough stuff fast enough.” It was always “are we making the right stuff?” That question hasn’t gotten easier just because the canvas is more responsive.

What the new toolset has done is raise the bar on execution quality across the board. This means the standard for differentiated thinking has risen in parallel. Design execution, in the sense of pixel-level production work, is increasingly table stakes. Strategic insight and understanding of the problem to be solved and audience it is being solved for (whether human, machine or both) is increasingly the differentiator and the critical role that designers must play now.

One thing we’ve committed to internally: our teams experiment with new tools as a standing practice, not an occasional initiative. The landscape is moving too fast for a “we’ll evaluate this in Q3” posture. If something new ships, someone on our team is using it within days. That culture of experimentation isn’t just about staying current; it’s how we learn to distinguish tools that deliver real impact from tools that are the current headline.

That said, we’d be remiss not to mention the cost reality. The commercial models around design tooling are shifting in ways that deserve conversation with clients. Platforms that once operated on straightforward per-seat pricing are moving toward consumption-based models. Figma’s move toward token-based pricing being a prominent example, and the cost of design tools have risen meaningfully. That’s not a reason to resist new tools. It is a reason to be intentional about which ones earn their place, and to factor tooling economics into budget planning rather than treating software as a fixed, invisible line item. The reality is that budgeting for tools that use AI will require a new way to think about it and tactical approaches to managing budgets may be required in the short term while the pricing models evolve.

Design-Driven vs. Product-Focused: This isn’t a dichotomy anymore.

For most of our industry’s recent history, there’s been a low-grade tension between design-led and product-led ways of working. Design wanted more seats at the table. Product wanted faster cycles. Internal organization structures often were getting in the way of the work. We see this design vs. product tension dissolving. Not because one side won, but because the conditions that generated it have changed.

Now with tools and process changing with AI, it is enabling the two disciplines to come together.

When design execution was slow and expensive, you had to choose: invest deeply in design exploration or move fast on product. Today, when the cost of generating and testing a direction has dropped significantly, that trade-off is less sharp. There’s less reason to treat design rigor and product velocity as opposing forces.

What we’re arriving at and what our best client work increasingly reflects is an integrated model: design and product thinking woven together from the earliest stages, not handed off in sequence. The question “what should we build?” and the question “how should it feel and function?” are being asked in the same room, at the same time, by people who can hold both.

Users Are Still the Center. Full Stop.

This one isn’t up for debate, but it’s worth saying out loud because the noise around AI, speed, and efficiency can quietly push it to the margins.

User-centered design isn’t a methodology. It’s a commitment. It means that when there’s tension between what’s easy to build and what actually serves people, you resolve it in favor of the person on the other end. That hasn’t changed. If anything, in an environment where now it is easier to generate large quantities of designs and code, the discipline required to keep users at the center and to resist the pull of “this is fast, ship it” has become more important, not less.

Where we’ve evolved is in how we center users. Research is faster and more continuous. Synthesis tools and ability of AI to analyze and summarize results have dramatically improved. It is now much easier to weave user insights throughout a project rather than front-loading it into a discovery phase that gets compressed when timelines tighten. The creation and usage of synthetic personas and journeys can help augment and accelerate programs. But the underlying commitment is the same as it was on day one. You need to nail the real business need and not just design a solution in search of a problem and the best way to get clarity on a business problem is to get insight from the actual users. The good news is now there are even more ways to get this through new tools, techniques and being smart about how and where to leverage AI.

Smaller High-Performance Teams are the Norm.

The team of ten that used to be the default model for a medium-sized project is increasingly a team of three or four who can move faster and accomplish more. That’s not a budget play. It’s a structural reality: when individuals carry broader capability, you need fewer handoffs, fewer coordination layers, and have fewer dependencies.

We look less for experts and more for directors who can own a wider surface area and wear multiple hats. Our team members can go broad across the lifecycle as well as deeper in one more than one area. For example, most of our designers are able to lead, design, produce and all have deep industry experience. While this is the way we have always built teams, we now see our clients moving away from the ‘factory models’ that enabled scale in favor of smaller higher performance, more versatile teams. While individual hires may come at a higher price tag, they usually can deliver at a higher velocity with strategic results.

That’s a different hiring challenge, and a different way of thinking about team composition on any given engagement. What does remain a reality is that while there was a time a few years ago when clients were looking to us to staff out projects with more junior resources for better leverage the model has actually flipped where more senior, experienced and multifaceted consultants are in demand.

Some roles are under real pressure. Production design that includes the work of implementing, adapting, and scaling visual assets at volume has been substantially affected by AI-assisted tooling. Tasks that once required dedicated resourcing can now be handled faster, often by the same person doing upstream design work. That’s not a forecast. It’s what we’re seeing in practice.

We don’t think this means production skills are worthless. We think it means that pure-production roles, defined narrowly, face a real commoditization curve. Practitioners need to think about moving beyond producing themselves to creating skills, prompts and toolkits to enable production. This is critical for the next generation of designers and that we all support them to become well rounded to fit into this emerging model.

Design research as a standalone function is under different but related pressure. The tools for synthesis, pattern recognition, and insight generation have improved significantly as mentioned earlier. That doesn’t eliminate the need for experienced research skills.  If anything, it surfaces the fact that the expertise layer (knowing what questions to ask, how to interpret ambiguous signals, what findings actually mean for a design direction based on experience in the industry) is where the value sits. But it does compress the headcount required to run a robust research practice as well as timeframes with which to conduct research activities and programs.

What’s not going away with AI: design strategy, systems thinking, facilitation, and the ability to translate between user needs, business goals, and technical reality. These skills have, if anything, as alluded to earlier, gotten more valuable as the pace of everything else has accelerated.

A Few Hot Takes We’ll Stand Behind

Design strategy is having its moment and it’s overdue. For years, strategy work in design was treated as a premium add-on. Now clients are asking for it first. When you can prototype anything quickly, the question of what to prototype becomes the highest-leverage decision in the process.

The generalist/specialist debate is over, and the hands-on directors won. The most valuable people on our teams aren’t generalists who know a little about everything or specialists who know everything about one thing. They’re practitioners who have gone deep in two or three adjacent areas and can move fluidly between them and bring industry knowledge to the table also.

“Design thinking” as a workshop format has run its course. The practice it pointed at, bringing genuine design rigor into problem-framing and strategy; is more relevant than ever. The two-day sticky-note sprint as a substitute for that rigor is not.

Smaller teams are a feature, not a compromise. Leaner, more integrated project teams move faster, communicate better, and often produce sharper work. We’d rather staff four people who each hold two competencies than eight people who each hold one. And if they have domain / industry expertise (which our team members do) even better.

Where We’re Going

Design is not a department, a deliverable, or a phase. It’s a way of working.  One that asks hard questions early, prototypes its way to clarity, and keeps the person on the other end of the experience in the room, even when they’re not in the room.

The tools will keep changing. The team structures will keep adapting. The specific skills that are scarce today will be table stakes in three years.

What won’t change is the underlying discipline: making things that work for people, in the real world. That’s the work. We’re still doing it.

We’d love to hear your thoughts on this. Please contact us to continue the discussion or discuss our take on approaches to solve your business problems.

NovusNorth is a design and product strategy firm. This piece reflects our current thinking — which, in keeping with everything above, we reserve the right to revise.