Reskilling throughout an organization’s workforce can help avoid an AI divide.GETTY IMAGES
Deploying artificial intelligence within an organization is not just a technology exercise. It’s also part of a work culture shift.
That’s why Punchcard Systems, a digital product studio based in Edmonton, has exposed its entire team to AI tools, trained everyone and given them the chance to experiment and fail.
“We have to be okay with learning and relearning,” says Sam Jenkins, the company’s co-founder and managing partner. “We’re all back at school when it comes to the potential of the tools that are hitting the market.”
Punchcard Systems builds and modernizes revenue-critical software platforms for clients in sectors such as health regulation and financial services. The company has nearly 50 full-time employees, and as part of its AI use policy it runs regular skill‑sharing sessions and it created an internal AI chat interface for staff access to models such as Cohere, ChatGPT and Anthropic.
“This was used by all roles, not just our technical folks,” Mr. Jenkins says.
His aim is to create a “psychologically safe environment” where his team can try new tools, make mistakes and learn without introducing risk to clients. By the end of this year, Punchcard aims to achieve AI competency for all staff.
Companies should be thinking about AI adoption from a HR context, says Suchit Ahuja, associate professor of Supply Chain and Business Technology Management at Concordia University, and co-director of the school’s Applied AI Institute. Unless employees get opportunities to reskill, he says an “AI divide” will perpetuate. One set of staff will know how to exploit AI for their job, and another set will “simply get left behind because they never took the leap.”
Decades ago, organizations had to decide whether they were Internet-driven. Now, Mr. Ahuja says, they need to decide whether they’re AI-driven, and what type of training and governance that need requires.
Mr. Jenkins calls the rapid pace of change “thrilling,” but he acknowledges it can feel disruptive. The company recently held a series of internal town halls for staff to share their feelings about the developments. “I feel vulnerable, too,” Mr. Jenkins says. “Change is difficult, and for many this is an overwhelming time.”
He adds that broader use of AI doesn’t necessarily spell layoffs, but the ratio of future hires may shift. For example, if a software developer can work faster with AI, Punchcard may need more product managers to oversee that output.
An early adopter of AI, Punchcard uses the technology to streamline its internal operations, such as making a first-pass review on its code. However, there is always a human on the other end to validate the work.
“The real skill today is not prompting ChatGPT. It’s knowing when not to trust the output. Without that, we get AI slop at scale,” Mr. Jenkins says.
AI, he points out, doesn’t just optimize existing workflows, it also exposes where they’re weak. “Most companies think that they’re very well structured when they’re really not. AI forces us to be operationally clear. It reveals how messy things actually are.”
Mr. Ahuja agrees, saying AI isn’t a “magic wand” that fixes fundamental business challenges. If a product is built on flawed data, he suggests AI will only “aggregate the pain points even more.” Or, as Mr. Jenkins says, “it makes bad decisions faster, just as much as the good ones.”
Punchcard clients increasingly expect tighter turnaround times. While AI can speed up parts of the work, Mr. Jenkins says it does not decide what products to build or which features to prioritize. “It’s accelerating execution but not solving for clarity.”
Companies gain an edge by understanding what AI can and cannot do best, and by having a culture of curiosity. That’s why Mr. Jenkins isn’t just looking for AI investments but for candidates who value adaptability and continuous learning. “Being in technology means that we can’t stand still. We have to be continuously adding to our personal encyclopedia.”