‘If employees are not using the tools, they are falling behind your competitors’: academic explains why HR may want to delay laptop refreshes and double-down on training
Canadian employers are facing a new kind of technology budgeting problem: the AI boom is making the basic hardware behind everyday workplace tools more expensive.
Memory manufacturers are shifting production away from consumer products toward AI data centres that “need massive amounts of memory to operate,” contributing to a major spike in RAM prices, according to the CBC.
Ajay Agrawal says this hardware crunch has broad implications for employers.
“Basically, it means almost anything with a computer in it will become more expensive,” says the Geoffrey Taber chair in entrepreneurship and innovation and professor of entrepreneurship at the University of Toronto’s Rotman School of Management.
“So laptops, cell phones, other devices – anything, basically with memory will become more expensive.”
Agrawal links those higher prices to a mismatch between capacity and demand in the global supply chain.
He notes that “we’re going through this transition where prices are bouncing around in ways that people didn't expect, because we didn't expect there to be such a surge in demand.”
According to Agrawal, the current environment calls for a rethink of established replacement cycles and for a sharper focus on where technology spending will actually move the productivity needle.
Stretch hardware lifecycles before buying new tech
The first adjustment Agrawal recommends is to think twice before replacing functioning hardware on an automatic schedule.
Many employers follow a three‑year cycle for laptops and other devices, but that may no longer be the most efficient use of tight budget – he suggests they stretch hardware lifecycles before buying new devices: “So rather than refreshing this year, if the machines are on a three-year cycle, extend that to four years.”
By lengthening device lifecycles, organizations may be able to free up funds for other technology priorities that are more directly tied to productivity gains, Agrawal says – particularly in AI‑enabled software.
He frames it as a choice between buying more machines on the same schedule and investing in cloud‑based tools that can make existing staff more productive.
“They can extend the life of those for an extra year and instead allocate those funds to cloud computing or software running on the cloud, to take advantage of these AI tools,” he says, adding that not all AI‑related software falls into the “experimental” bucket as some tools are already delivering measurable gains.
“A lot of these AI things are ‘nice to have,’ but some of them are a lot more than nice to have,” he says.
“They are very significant productivity-enhancing tools; in other words, they help their employees to be able to produce 20 to 30 percent more than what they were producing without the tools.”
Cloud services versus on‑premise systems
Alongside the hardware-software trade‑off, Agrawal points to a structural shift in how employers keep technology current. On‑premise systems that require periodic major upgrades can be difficult to update in an environment where capabilities are changing on an almost monthly basis.
Where business‑critical constraints allow, Agrawal says organizations should consider relying more on software‑as‑a‑service (SaaS) for non‑core functions: “Anything ... that's not critical to the business, that they can use cloud services [for] as opposed to operating on premise,” he says.
“Because the software as a service … those things are being upgraded so frequently, [it’s] just a much more efficient way to stay current than upgrading internal systems.”
Training employees to keep pace with AI tools
Agrawal notes that most office workers today are used to software that evolves slowly, with long periods between major interface or feature changes. That assumption no longer holds for many AI‑enabled tools, and he stresses the importance of a focus on skills rather than hardware or system architecture.
“Most people know how to use Microsoft Word or Google Docs … they learn them, and they're pretty much good to go for a long period of time,” he explains.
“Right now, these AI tools are advancing so fast that if you learned how to do something two months ago, chances are that there's a significantly better software solution today, and the people who learn that are 10, 20, 30 percent more productive than the people who aren’t. There's a very big premium right now to training people.”
Basically, money spent on licences and infrastructure may not translate into better outcomes if employees don’t know how to use them, he explains. In this context, training is not an optional add‑on but part of the core investment in AI‑enabled systems.
“There are now AI tools in every category, so if you're in a company where your employees are not using the tools, then they are falling behind your competitors,” Agrawal says.
“This is the year to invest in upgrading skills to keep up with the software.”
