Grace Clarke is the founder of GGC Consulting, redefining how brands grow by building community-powered flywheels, not funnels, growing brands like Graza, Madewell, Jones Road, Estee Lauder, Target, Google.
She is also the founder of GraceAI, a marketing strategy and calendar generator (basically an AI clone of her brain). She also runs the Gen Z Board of Directors, an insights network of young people, is an artist, and splits her time between New York and Paris.
For Grace, the conversation around AI has moved far beyond asking a chatbot to write an email or summarize a document. The real opportunity is to build systems that work quietly in the background, taking on repetitive processes, surfacing what matters, and giving leaders more space to focus on the work that requires uniquely human judgment.
“I am not an advocate of removing humans from the loop,” Grace said. “The promise is that AI gives us more room to think critically and do the things that only we can do in the world and with each other.”
Grace sees the next phase of AI as a shift from using technology on demand to allowing it to become an active part of how we work. Rather than opening ChatGPT every time there is a task to complete, leaders can begin creating autonomous systems that monitor information, move work forward, and return to them only when human input is actually needed.
She described these systems as a kind of “shadow” of the user: an AI with a defined job that can operate in the background, connected to the tools and information it needs to complete that job. That might mean monitoring a CRM, organizing information from meetings, preparing a proposal, or identifying a problem before it becomes a larger one.
The important distinction, she emphasized, is that building an agent does not necessarily mean building something technically complex. In many cases, an agent is simply a collection of skills and instructions that work together. One skill can even call another, creating a workflow that would otherwise require a person to manually move from one platform, document, or task to another.
For leaders who feel overwhelmed by the pace of AI, Grace's message was intentionally practical: you do not need to understand every technical layer underneath these systems. You need to understand the problem you want solved.
The hardest part of building an AI agent may not be the technology at all. It is deciding what is actually worth building.
Grace encouraged the group to resist the pressure to automate everything at once. Instead, start with the moments in your work when you find yourself thinking, It would be so great if AI could just do this.
Her advice is to pick one problem. AI can help identify possibilities based on the context it already has about your work, but the human still needs to decide which one is worth pursuing. Once that decision is made, block time on the calendar to work on it. Even an hour can be enough to begin mapping the process.
Grace also recommended asking AI to visualize the workflow before building it. Mapping each step makes it easier to catch missing information, unnecessary steps, or places where the system could go wrong. It also changes the process from something that feels highly technical into something much more familiar: simply describing how you already work.
“The blank is really all that we need to get started,” she said, encouraging leaders to begin with the task they wish AI could take off their plate.
One of Grace's simplest examples is what she calls a daily sweep. Instead of beginning each morning by sorting through emails, messages, records, and updates, an AI system can review connected platforms at the end of the day, update records, assign action items, and identify anything that requires attention.
The value is not simply saving the time it takes to update a CRM. The larger benefit is cognitive relief. The system can surface the fact that a deal appears to be moving, that inventory may be running low, or that something needs to be addressed before it becomes urgent. Rather than asking a leader to remember every detail, the system brings the important decisions back to them.
Grace demonstrated how the same principle can be applied to relationships. A client relationship agent, for example, can track a person from an initial meeting through follow-up, research, proposals, and next steps. When a relationship reaches a particular stage, the system can trigger another skill to develop a proposal or brief, pull in relevant context, apply a brand's voice and working guidelines, and prepare a finished document.
But there is an important human checkpoint. Grace builds what she calls an “exit condition” into these systems. In her example, the AI can prepare the proposal, but it cannot send anything without her approval. The system handles the work; she remains the final decision-maker.
That same thinking can be applied to business performance. Grace described “watchdog” agents that monitor a defined set of metrics at regular intervals and alert a team when something crosses a predetermined threshold. Instead of discovering a problem a week later, a business can be alerted on day one, along with a recommendation for what to investigate or do next.
For Grace, this is where AI becomes more interesting than a traditional dashboard. A dashboard tells you what happened. A more sophisticated AI system can look across business data, customer sentiment, market trends, competitor activity, and other signals, then help force-rank the actions that could have the greatest impact.
Her philosophy is simple: no more dashboards without decision-making support.
Getting AI into a business is not simply a technology challenge. It is a leadership challenge.
Grace emphasized that one of the most important factors in team adoption is that the leader goes first. Before asking employees to change how they work, leaders need to experiment themselves, understand the tools, and develop enough confidence to articulate where AI can genuinely add value.
“The leader goes first,” Grace said.
That means organizations need more than access to AI tools. They need thoughtful policies around safety, permissions, public-facing use, employee documentation, and training. Grace also recommended identifying internal champions who can help teams adopt the technology thoughtfully rather than treating AI as another software rollout.
The goal is not simply to give everyone an AI account. It is to create a shared understanding of how these tools should be used, where they can create leverage, and where human judgment must remain firmly in the loop.
Perhaps the most important point Grace made was also the least technical.
As AI becomes increasingly capable of taking work off our plates, leaders need to be careful about what they give away. Automation can create extraordinary amounts of cognitive relief, but there is a difference between removing administrative friction and outsourcing our ability to think.
Grace warned against what she called “cognitive atrophy” and “lazy prompting.” If AI makes our lives easier, the time it creates should not simply be filled with more work. It can create room for critical thinking, creativity, relationships, hobbies, time offline, and the parts of leadership that cannot be automated.
Her recommendation is to keep thinking alongside the technology. Rather than asking AI to make every decision, start with your own point of view: Here is what I think I should do. What do you think?
That distinction is important. The best AI systems are not necessarily the ones that remove humans from the process. They are the ones that make humans better at the parts of the process that matter most.
Grace's final message was less about keeping up with every new AI product and more about developing the confidence to build something useful.
The pace of innovation can create constant FOMO, but trying to experiment with ten ideas at once can become its own form of paralysis. Grace suggested choosing one thing and giving yourself permission to ignore the rest for 72 hours. The act of building one useful system, she argued, will teach you more than collecting a list of possibilities ever could.
AI is evolving quickly, but the leadership question is becoming clearer: What work should technology take off your plate, and what work should it give back to you?
For Grace, the answer is not to work less thoughtfully. It is to create more room to think, decide, create, and lead.
