• by Laurie Sullivan , Staff Writer @lauriesullivan, Yesterday

Paul Woolmington, chief executive officer (CEO) at Canvas Worldwide, encourages employees to create their own AI-based projects and tools. He has been working on a few of his own. One he calls Bullfighter, although he is not quite sold on the name.

The tool analyzes articles and documents. Woolmington has been testing it on company reports, feeding documents into AI to identify cliches and jargon.

“The more flowery language CEOs use, the less likely they are to perform in the S&P 500,” he says.

“I hate words like pivot and synergy -- nebulous words that mean nothing,” he says, admitting it’s a hobby project. He also dislikes acronyms, but the fun part for him is eradicating the jargon. “Over pivoting, what does that mean? It means your business is screwed.”

The word "creative" has taken on new meaning -- not creative in ad content, but creative in thinking. Woolmington has made checklists and built a grid for different prompts for ChatGPT and other AI chatbots like Perplexity. This technology understands ways to get the most from applications.

Until now, he never thought about creating a grid with economic factors that might influence a company’s business, and discovering them just by feeding the words as prompts into AI models.

Woolmington says his next favorite app will likely become the one he makes himself, which would have seemed unlikely just a year ago. Software has been something companies and individuals used, but not had the ability to build themselves. Not until AI came along.

Woolmington's curiosity led the company deeper into data-based AI. He detailed Canvas's approach to AI and the development of the AI-driven tool such as Ella, a privacy=compliant ChatGPT-like AI platform that anyone at the agency can work with, as well as partnerships with Microsoft and Chalice.

Campaigns are becoming personal and leaving personalization behind. AI is needed to complement human intelligence rather than replace it.

Chalice is known as an AI company for connecting buyers with quality media across the web, building custom algorithms, and more to improve programmatic media buying.

“We’re not just adapting to AI, but rather re-looking at everything we do and rewriting the rules of how brands connect with consumers at a fundamental level,” he said during the conversation. “There are three parts, and the trinity of convergence is upon us.”

Canvas has been working with Chalice to create problem-solving tools driven by AI that understand in real-time where products are available. The tool identifies product availability across the United States, and then shifts media budgets to that region.

“Why advertise in certain geographies if we don’t have the models and the products?” he said.

The company is rolling out the tool to all product categories that have similar challenges. Woolmington described the tool as a way to use AI for media-buying intelligence. 

“Predictive intelligence is no longer optional, and the notion of speed and efficiency is part of tech maturity,” he said. “Every channel is now digital.”

The future is AI as a creative copilot, not a human replacement. Predictive modeling, not just retrospective measurement, and creative intelligence that can adapt in real-time to fill consumer expectations, he said. Then enterprise intelligence that integrates across all workflows.

The goal has become personal media, not personalization, as well as media intelligence to optimize insights, planning and buying.

“I liken that to a Formula One Ferrari because the new engines we’re building are quite sophisticated machines,” he said. “We need lots of Lewis Hamiltons [a British race car driver] trained to drive them because you cannot just step in.”