Charlene Li on AI, Architecture, and the Future of Work

Charlene Li is a New York Times bestselling author, speaker, and strategist who has spent decades helping leaders navigate disruptive technology and organizational change. In this episode, the Arcfluence Team sits down with Charlene for a grounded conversation about AI, architecture, and what it takes for firms to adapt with clarity instead of fear.

Charlene frames AI not as a magic solution or a replacement for human expertise, but as a leadership, culture, and decision-making challenge. The conversation explores how architecture and design firms can use AI to reduce friction, improve client communication, rethink workflows, support talent development, and create new value in the earliest stages of a project.

Rather than chasing every new tool, Charlene argues that firms should start with the problems they need to solve, build AI fluency across their teams, and create governance that helps people move faster without losing control. The result is a practical discussion about experimentation, judgment, adaptability, and the human expertise that still sits at the center of good design.

It starts with understanding what you actually mean

Most projects do not begin with a clean brief. They begin with a feeling. A space that no longer works the way it used to. A layout that fights you. A sense that something needs to change, even if it is hard to put into words.

That is not a gap in the process. It is the beginning of it. Architecture has always been about turning half-formed intent into something that can actually be built. And that translation, from how a space feels to what it actually needs, is exactly where Charlene sees AI becoming genuinely useful.

When you tell an architect a room feels closed off, or a lobby does not feel welcoming, or a space technically works but does not feel like you, those are real design inputs. They just need interpretation. AI can sit in that gap alongside Pinterest boards, reference images, and early conversations, helping translate in both directions: turning what you feel into language that reflects intent, and turning technical decisions into something more understandable and actionable.

It does not make the judgment for anyone. Used well, it helps make the conversation around those judgments clearer, earlier.

A way to explore directions, not an answer machine

The least interesting way to use AI is to ask it for the answer. The more useful way is to use it to see further and map out the directions a project could take while those directions are still easy to change.

Danny put it this way:

Early in a project is when many of the most important decisions are made: how a site is used, how a space wants to be lived in, which tradeoffs are worth making, and which ideas should be left behind. Used well, AI can help a design team explore more of that territory with you before anyone commits to a path.

That is less about producing faster and more about thinking wider, together, while it is still possible to change your mind. Charlene’s own habit makes the point: she ends her prompts by asking the tool to ask her questions first. That turns AI from something that hands back an answer into something you can reason with.

From being shown a design to shaping it

For a long time, parts of the design process worked like a series of reveals. The architect goes away, does the work, comes back, and the client reacts. That can work. But it can also create distance between the thinking behind a design and the client’s understanding of it.

Charlene pointed to a shift that larger firms are already naming: the architect moving from presenter to conductor, coordinating options, constraints, and stories into a shared direction. That changes the role the client can play, too.

Nick described the same challenge from inside the practice: the hardest part is often not the design itself, but the communication around it.

When that communication improves (options can be shown, discussed, and adjusted while a project is still taking shape), the back-and-forth becomes richer instead of simply more expensive. The client is not just signing off on a finished idea. They are helping form it.

The thinking becomes the most valuable part

The part of Charlene’s argument with the longest reach was not really about efficiency. It was about where the value in a project actually lives.

For years, the early strategic work (siting, zoning, programming, feasibility, massing, and the thinking that happens before anything looks like a building) was often treated as the cost of winning the job. Something a firm gave away, or folded quietly into the larger fee. AI changes that math.

When a team can run hundreds of zoning, massing, or planning scenarios and return with a real recommendation, that early thinking stops being throat-clearing. It becomes one of the most valuable parts of the work.

The drawings still matter. The built result matters most. But the quality of the thinking that happens before either of them is where a project is often won or lost.

Honest about what it cannot do

None of this works without expertise in the room. AI is most useful to people who already understand the work, people who can tell which thread is worth pulling and which one leads nowhere. Danny’s framing is that AI augments someone who already has a background. Drop a person into a subject they know nothing about, and they may not know which thread to pick up in the first place.

AI can produce something that looks convincing and still miss the point. On its own, it does not carry constructibility, zoning reality, accessibility, long-term durability, or the lived experience of a space. Those are exactly the things a good architect holds in mind throughout the process, and exactly the things worth being honest that a tool, by itself, will not replace.

That is the line between a tool in the mix and a replacement for the people doing the work. It is firmly the former.

No magic wand

Charlene calls herself a realist optimist, and it fits the conversation. There is real potential here, and real work in using it well.

At Arcfluence, we tend to think about design as something that grows through conversation. Nothing about AI changes that. Used honestly, it gives that conversation more range: a faster way to understand what you mean, more directions to weigh together, and a clearer path from a vague feeling to a shared decision.

That is the part worth caring about. Not the technology for its own sake, but a better way to understand each other and, in the end, better places to live, work, and invest.