Tricia Wang

Most companies are training people to be better users of AI, and that goal is set too low. A user takes whatever the machine returns. The people getting remarkable results do something else entirely. They steer it, push back on it, and bend it toward what they actually need. Tricia Wang calls them Shapers, and she has spent years in the field with the ones already doing it, from artists and immigrants to call-center teams to airline pilots to inside large enterprises.

She found that Shapers share a skillset anyone can learn, and she captures it in a framework called VIDA. Verify the human at the source. Interpret what the output means. Decide what to do. Stay accountable for the result. The order reflects the whole idea chain that accountability rests on a decision, the decision rests on interpretation, and interpretation rests on having verified something real to begin with. Her talk documents various winning moves of enterprises that have adopted AI strategies that position their employees as Shaper where people and machines in an ecosystem where each is doing what they do best.

Audiences leave able to recognize Shaper behavior on their own teams, apply the four VIDA steps to live work the next morning, and design strategy and training that grows judgment rather than dependence.

Every board wants agents in production. Far fewer leaders can answer the question that should come first: can you prove what your agents actually did? An agent does not sit still behind a firewall. It crosses clouds, calls outside services, and moves money on its own, with no perimeter left to defend. Security spent decades guarding a wall, and agents walk through walls by design. They are also only as trustworthy as the data feeding them, which is where most of the risk is concentrated in companies.

When Air Canada's chatbot gave a grieving customer a refund policy that did not exist, a tribunal made the airline honor it. One agent, one confident wrong answer, and the company owned the fallout. That is the world leaders are deploying into, and seeing these shifts arrive early is what Tricia Wang has done her whole career. As CEO of Advanced AI Society, she is building the security ecosystem the industry is still debating. She walks leaders through the change already reshaping the C-suite, from controlling where agents go to proving what they did. Security used to be the office that said “no” and appeared only when something broke. With agents, security becomes the function that decides how much a company can safely turn loose.

Executives leave able to articulate to their board and team why security is a critical part of a winning AI strategy. After this talk, AI security safety are nice to have slogans for an ethical or responsible deployment, but the foundations of the entire plan.

The promise of AI is that enough data and enough automation will make good decisions happen on their own. It rarely plays out that way. Every model runs on a flattened version of reality, and the flattening always drops something. Leaders who can name what got dropped make better calls than the ones who trust the screen, and teaching them to see it has been Tricia Wang's work for more than a decade. She coined Thick Data and Quantification Bias, the vocabulary executives now reach for when they argue about their own numbers.

Both ideas matter more once AI is in the room. Big data is the trace people leave behind. Thick Data is the meaning they carry. A model reads the trace beautifully and stays blind to the meaning, and that blind spot is where human judgment does its real work. Wang has watched what happens when companies forget this. Her talk shares stories with actionable lessons on how to leverage LLMs and dashboards in a way that avoids outsourcing the human judgements to them. They also leave with language to describe something they have sensed for years and could never quite name: the winning move for humans as AI is increasingly automating more human activities.

Leaders hear that AI will change everything and almost never hear how it works. That gap is where the worry lives and where the expensive mistakes get made. Tricia Wang closes it. For a decade, companies paid her firm to repair the data foundations under their AI ambitions, so she can open the hood and explain what is really happening, with none of the hype and none of the doom.

She traces the line from big data to today's models in plain language, and she is blunt about why so many AI projects stall. The model is usually fine. The data underneath it was never cleaned, never governed, never ready. From there she gives leaders a working read on what is genuinely usable now, what is still oversold, and what to do about it this quarter. People who walked in wary are able to walk out able to explain AI to their own teams, question a vendor without bluffing, and tell a real opportunity from a shiny one. It is the rare AI briefing that lowers the temperature in the room and raises the competence.