Three parts in, I've argued that taste alone isn't sufficient defense against what's coming, that the shape of valuable design work is changing faster than hiring practices are adapting, and that the window to respond is shorter than most organizations are behaving as if it is.
The natural question is: so what do you actually do?
Not in five years. Not when the trajectory of AI becomes clearer. Not when your organization finally gives you a mandate to change. Now — with the information you have, in the organization you're actually running, against the constraints that are genuinely in front of you.
The answer is not to predict where AI ends up. Nobody knows. The organizations betting everything on a specific AI future are taking a risk that isn't necessary. The ones paralyzed by uncertainty are taking a different risk that's just as real — and compounding it quietly while they wait.
The right frame is simpler and more durable than either of those positions: plan for the horizon you can see.
The horizon you can see is roughly 24 months. Within that window, the broad shape of what's happening is clear enough to plan against, even if the specifics aren't.
AI is going to continue expanding the capability and volume of what design tools can produce. The execution layer of design work is going to continue being automated at pace. The designers who are learning to operate in the curation-orchestration-curation-presentation model are going to become increasingly valuable relative to those who haven't. The gap between organizations that have built genuine AI fluency into their workflows and those that haven't is going to widen.
Beyond that window, the picture becomes genuinely uncertain. The long-term endpoint — AI systems communicating with each other, humans as bottleneck, the loop closing without human involvement — is a real trajectory. But it is not the horizon you are managing to right now. Planning for a horizon you can't see yet, at the expense of the one you can, is how organizations get caught flat-footed by the transition that's actually in front of them.
This is a discipline as much as a strategy. It requires resisting two temptations simultaneously: the temptation to extrapolate so far into the future that the present feels unmanageable, and the temptation to stay so focused on the present that the future arrives as a surprise.
The leaders who navigate this well are not the ones with the most sophisticated AI predictions. They are the ones who hold the visible horizon clearly, make deliberate moves toward it, and build enough organizational flexibility to adapt when the next horizon comes into view.
This is the part where the series has to be honest about what it is and isn't. Four essays can identify the territory. They cannot tell you exactly what to do inside your specific organization, with your specific team, against your specific constraints. Context matters enormously. The right moves for a five-person design team at a Series B startup are not the same as the right moves for a fifty-person design organization inside a global enterprise.
What follows are the moves that hold up across contexts — not because they are universal solutions, but because they address the underlying dynamics that are universal.
Hire for earned taste and adaptability. Both.
The temptation in a moment of uncertainty is to hire for one or the other — either the experienced, opinionated designer with deep judgment, or the adaptable generalist who picks up new tools quickly. The profile that actually serves you in the next 24 months is both: a designer with genuine experiential judgment who has also demonstrated the intellectual flexibility to operate in new modalities rather than waiting for them to stabilize.
This person is not easy to find. They are not always the most impressive candidate in a traditional interview process. They may not have the most polished portfolio. The signals you're looking for — earned taste that is precise and communicable, combined with demonstrated adaptability to new tools and workflows — require a different evaluation framework than most hiring processes currently provide.
The investment in finding this profile is significant. The return is a team that is positioned for the transition rather than running behind it.
Treat AI fluency as a team capability, not an individual hobby.
The current state in most design organizations is that AI engagement is self-directed. The designers who are curious explore the tools. The ones who aren't don't. The result is a growing capability gap within the team — between the designers building genuine fluency and the ones who aren't — that most design leaders don't have a clear picture of.
Closing this gap requires treating AI fluency as an organizational capability rather than a personal development choice. That means building it into how the work gets done, not alongside it. It means creating explicit expectations about what fluency looks like at different levels of the team and what the pathway to developing it is. It means running projects in ways that require engagement with the new tools rather than making that engagement optional.
None of this is complicated in principle. It requires leadership will more than organizational resources.
Update what you're evaluating in your hiring process.
If your interview process is still primarily designed to surface execution-layer craft skills — portfolio quality, Figma proficiency, process documentation — you are evaluating for the old job. The candidates who will perform best in the AI-augmented environment are often not the ones who look best in that process.
Updating the process means adding evaluation moments specifically designed to surface the capabilities that actually matter now: curatorial judgment under pressure, comfort with ambiguity and orchestration thinking, the ability to communicate AI-augmented work with authority and conviction. None of these require exotic new interview formats. They require thinking clearly about what you're actually trying to learn and designing the process around that, rather than around what's most familiar.
Run at least one project differently.
The most direct way to understand what the new model of design work actually requires inside your organization is to run a project using it. Not as a side experiment. As real work, with real stakes, evaluated honestly against real outcomes.
This means approaching at least one upcoming project with an AI-augmented workflow from the start — defining the curation layer upfront, building orchestration into how the work is structured, evaluating the output with the same rigor you would apply to any deliverable. What you learn from that project will be more useful than anything a conference or an article can tell you, because it will be specific to your organization, your team, your constraints.
The learning compounds. The first project teaches you what the second project should do differently. By the time you've run half a dozen projects this way, you have built genuine organizational knowledge about how the new model works in your specific context.
Make the organizational case before you're asked to.
The design leaders who will have genuine organizational support when this transition accelerates are the ones who built that case proactively — before the pressure to change became obvious, before the budget conversations forced the issue, before the talent gap announced itself.
Making the case proactively means being able to articulate, in business terms, what the transition requires and what the cost of not making it is. It means presenting a clear-eyed view of where the team is now, where it needs to be, and what the pathway between those two points looks like. It means treating the organizational change as a design problem — one that requires understanding your stakeholders, their concerns, and what it takes to move them — rather than a communication problem.
This is harder than it sounds when the urgency isn't yet visible to the people around you. But the leaders who have done this work before the window closes will find that they have organizational credibility and support when it matters most.
At some point beyond the visible horizon, the dynamics described in this series reach their logical conclusion. The loop closes. AI systems interact with each other without meaningful human involvement in the middle. The curatorial and presentation functions that are irreducibly human right now become less so, as social norms shift and organizational trust in AI judgment accumulates.
That endpoint is real. It is not cause for despair — it is cause for honesty about the trajectory and what it implies for how designers and design leaders think about their roles over the long arc of their careers, not just the next two years.
But it is not the planning horizon. And treating it as if it were — either by ignoring it or by letting it paralyze the decisions that are actually in front of you — serves nobody well.
The job of a design leader in this moment is to hold two things simultaneously: an honest view of where this is going, and a clear-eyed focus on the decisions that are actually within reach. To plan for the horizon you can see, while building enough organizational flexibility to adapt when the next horizon comes into view.
That has always been what leadership in uncertain environments requires. The uncertainty is new. The requirement isn't.
After Taste started as a provocation: taste is necessary but not sufficient, and the design community is hiding behind it.
It ended here: with a practical frame for what design leaders can actually do in the window that's in front of them.
The four parts connect. The taste argument failing doesn't mean taste doesn't matter — it means taste alone isn't a strategy. The shape of valuable design work changing doesn't mean designers aren't valuable — it means the nature of that value is shifting and the organizations that recognize that shift earliest will have the most time to adapt. The 24-month window being real doesn't mean the situation is dire — it means the cost of inaction is higher than the cost of imperfect action, and the leaders who internalize that will move first.
Plan for the horizon you can see. Move before you have complete information. Stay honest about what you know and what you don't.
That's always been the job. The tools change. The discipline doesn't.