Shelby AUSTIN Cooper’99 is no stranger to global controversy. In fact, her business is at the centre of one of the great debate topics of our time—artificial intelligence. But that’s fine with her. “Half the world thinks this is all hype,” she says. “And half the world thinks we’re on our way to a scary artificial general intelligence kind of moment. The truth is probably somewhere in the middle.” And, as she will explain, her part in it really involves baking chocolate chip cookies.
With an infectious enthusiasm, leavened with a disarming self-deprecation, she will tell you about the importance of what her Toronto-based tech company, Arteria AI, does through streamlining document flow in large enterprises, and how it will improve lives as it continues to grow. Which it almost surely will, given her track record of the past decade or so, jumping from young lawyer to startup founder to high-powered global enterprise exec to her current status as CEO of this successful spinoff venture.
She’ll also be the first to tell you that most people think working with documents is “the most boring thing on the planet.” But listen to her account of how she named the company. “We believe that documentation is, really, this beautiful ‘aha’ layer in organizations. We think it’s like the fabric. Nothing gets done without it. In a lot of cases, the contracts are the lifeblood of an organization. So we were looking for something that represented, in a visceral way, being the core—like an artery.”
And then there’s the other half of the name: AI. Artificial intelligence is critical to Arteria’s product—its “digital documentation infrastructure”—which it offers to its large, mainly financial, clients, including investment banking giants Goldman Sachs and Citi. It can automate certain operations across the enterprise, from securities trading to mortgage lending, which inevitably require a vast number of steps—read “documents”—that are often highly repetitive. People may have to retype information, or rubber-stamp simple decisions stage by stage. Arteria helps eliminate a lot of the box-ticking, bringing together and analyzing what’s known as “unstructured data”—collections of everything from emails and texts to standalone files—so that the humans involved gain more insights into the process while having much less to do. The result, she says: “As people do their jobs, they’re able to do them in a smarter way.”
What Shelby says she especially loves is the challenge in what she does. “I’m never obsessed with the solution, but I’m obsessed with the problem. And then I try to find people who are smart enough to solve bits and pieces of it. I’m the least sporty person on the planet, but I’m effectively there to be the coach of the team. That’s the job I love doing.”
Her high-energy year-end letters to staff even end with “Live long and prosper.” Yes, she’s a bit of a Trekkie, from the Jean-Luc Picard years (Star Trek: The Next Generation) and she can do the split-fingers salute. “There’s a small part of me that’s very, very nerdy, and letting it out is one of my favourite activities,” she says, laughing. So is she a techie? Does she code? “I did have a summer job doing coding when I was little. And I did learn a number of things in my basement, but to claim I’m technical would be laughable to my team. What those experiences helped me realize, though, is that I was technically literate enough to understand that there are solutions to the problem.”
Above all, she no longer thinks of herself as the legal expert she started out as. “It’s so funny because being a lawyer is one of those professions where people really develop a sense of identity around it, and I just lost that, probably a decade ago. My career took such a left turn.”
In fact, much of her time at Branksome—from SK to Grade 13—was built around the law. “I was a real humanities junkie at the time because I spent my whole early life thinking I would go to law school and be a lawyer.” But what she got from her teachers remains very valuable. “What they taught me to do is think critically, challenge the status quo, ask questions, get answers, ask more questions. And teaching girls to be leaders—why couldn’t it be me, even though I was never first in class in anything? Why couldn’t it be me if I worked hard enough and tried hard enough and gave more of myself to the problem? And then, you know, perhaps I had some gifts to give as well.”
at Branksome Hall in September 2023. Photo: Christian Peterson.
She went on to display those gifts through undergrad at American University in Washington, D.C., law school at Western and a budding career as a partner and litigator at Davies, a major Toronto firm. But she became fascinated with, yes, documents, and decided in 2010 to make the jump to starting a company, ATD Legal Services, that could handle them better through then-existing technology.
Which was hardly a smooth process, in her telling. “When I quit my job, I rented several thousand square feet, sat by myself with no support, and bought, like, 50 computers. It was completely bananas. Because I had no idea what I was doing—and that’s probably an understatement. I didn’t know how to run a company. I didn’t know how to get a client. I knew nothing!”
Did she have financial backing? “You know, I took money from my bank account. That company was bootstrapped from end to end. It was a crazy ride. But looking back, it was beautiful. And it was such a good learning experience.”
It was also so successful that four years later, Deloitte Canada, part of the global professional services firm, acquired it and installed her as a senior executive, eventually responsible for its acquisitions and leader of its Omnia AI team, the national artificial intelligence practice.
Then, in 2020, she and Deloitte agreed that a part of Omnia should be spun out as its own enterprise, with her at the top. Deloitte gave Arteria AI full backing and remains a warm supporter, but eventually unwound its interest as the company found other venture-capital backing and some powerful clients.
Now she is embroiled in AI, including all those controversies. “We are using very powerful tools that are real and can be very useful,” she says. “That doesn’t absolve us of being careful. But the world is changing so rapidly and a lot of what we’re hearing about is leaked research papers, not necessarily what’s going to be there in practice. So we have to be careful about what we rely on.”
Cue the cookies. Shelby points out that much of the hue and cry revolves around ChatGPT, the generative AI system that can create content, which does play a role in her system. But artificial intelligence involves many more systems and tools than that. “ChatGPT is one form of Large Language Model,” she says. “So, a subset of a subset of a subset that has some really useful things, but it’s not the only kind of model we use.
“What I would say is this: if ChatGPT and all of the Large Language Models are chocolate chips—and they are makers of delicious chocolate chips—I am not a chocolate chip maker. I am a cookie baker. And so my job is to bake the most delicious chocolate chips into my cookies, and then serve my cookies to whomever I can. I am a user of chocolate chips—sometimes I even add special things to make them more delicious—and then I add them to a much larger sort of stack.”
Her ultimate concoction, of course, automates a great deal of repetitive work conducted by, as she says, “thousands” of people working for her clients. To which some will ask: Will those workers be able to keep their jobs? Shelby is both candid and humane about that. Many of the repetitive tasks are gone for good, she says, though often it is “just 10 per cent of their job.” But she adds: “I’ve been very vocal that we have to be clear about what we’re offering the people whose jobs are these repetitive tasks. I think it’s not intellectually honest if you don’t acknowledge that some may lose their jobs.” Companies must look at retraining or other measures to take care of those displaced. More broadly, she says, “I think 100 years of economic history show us that AI will create more jobs than it will destroy.”
On her LinkedIn account this year, Shelby posted as “required reading” an influential essay by Marc Andreessen, one of the venerated pioneers of the internet, titled “Why AI Will Save the World.” While acknowledging some risks, he asserts: “AI is quite possibly the most important—and best—thing our civilization has ever created, certainly on par with electricity and microchips, and probably beyond those.”
“I’ve been very vocal that we have to be clear about what we’re offering the people whose jobs are these repetitive tasks. I think it’s not intellectually honest if you don’t acknowledge that some may lose their jobs.”
Shelby voices similar views, and has publicly urged Canada to get much more deeply into AI businesses. “For Canada to slow down and be the most conservative in its view of where this all goes will do nothing but hurt us,” she says. “I’d like to see young women bake more cookies—except in a less domestic kind of family!—by building awesome companies. I’d like to see more women getting startup seed money, which for them is less than two per cent of the global take right now.” She is an enthusiastic supporter of Branksome’s new Innovation Centre and Studio Theatre and a member of the Make Way Campaign’s Innovation Council. “It’s really spectacular. I can’t believe that not only are we going to enable this experience for our girls, but we’re also going to enable other schools to come and learn within that environment.”
There are a lot of lucrative AI niches for budding entrepreneurs to fill, she notes. In the future, “you will have an assistant to help you in every aspect of your life. And honestly, my hand on heart, I believe that leads to a better world.”
And yet, as someone who with husband Graeme Cooper has two kids—Simone, 10, who’s at Branksome, and Stafford, 8, at Upper Canada College—she is remarkably cautious about AI in her own life. She won’t use data-gathering gadgets like Alexa or Siri, and she has especially strong views about toys. “I don’t want a stuffed dog talking to my kids and putting them to bed. I don’t think kids are in a position where they can materially appreciate that it’s not an emotive human. I really think we have to be very careful.”
Careful, she feels, in creating a much better world.
Featured image by Jeff Kirk.
