TORONTO — As AI upends the software business model, some investors are looking to deep tech as a hedge against mounting uncertainty. 

“Every software VC is now all of a sudden also investing in deep tech,” said Jenny Yang, a Vancouver-based partner at early-stage deep-tech firm Risc Capital. The sector, overshadowed by software startups, is suddenly garnering lots of attention despite concerns about its speed, cost and uncertainty.

The growing interest comes as venture investors grapple with how AI is upending many of the software companies that dominate their portfolios. Software firms have traditionally derived their value from engineering talent and code. Recurring subscription revenue made cash flow stable and growth easier to forecast. New AI coding tools are eating away at some of those advantages by making it dramatically faster and cheaper to write software, with customers also less willing to pay for it.

That has pushed investors toward businesses they believe will be harder for AI to replicate. In deep tech, those advantages can come from manufacturing expertise, patent-protected scientific breakthroughs or proprietary data.

Neil Peet, partner at Toronto-based GreenSky Ventures, said his team is rethinking its portfolio to focus more on deep tech. The firm, which was founded in 2010 and is currently raising its seventh fund, has typically invested about 50 per cent of its capital in software and 50 per cent in deep tech. Now, it’s putting 60 per cent behind deep tech and 40 per cent on software.

Peet said GreenSky is also re-evaluating what kinds of software companies it backs, moving away from those that rely on the traditional software-as-as-service (SaaS) subscription model. Instead, the firm is placing more emphasis on whether a company’s technology is novel, difficult to reproduce and likely to retain value through intellectual property, for example, even if the company itself fails. 

“What we’re trying to do is to say this technology is actually really innovative,” Peet said.

Yang started to get uneasy about software startups in 2024 when it became clear how quickly AI was improving. Soon after, she joined Risc Capital, where she’s fully focused on deep tech. In her current role, she gets lots of calls from software investors seeking advice on how to approach deep-tech deals. “A lot of them are trying to figure out how they play in an area that has a bit more defensibility,” Yang said. 

She added that advances in AI have also made it more attractive to invest in breakthrough technologies, by helping speed up research and reducing the costs associated with it. Still, she said, it’s not easy to move from software to deep-tech deal making.

Investors entering the world of deep tech need to adjust their expectations around risk and return timelines, Yang said. Unlike software startups, which can quickly test and tweak products and change direction with relative ease, deep-tech companies often spend years developing physical technologies before generating meaningful revenue. Due diligence requires investors to have specific expertise in the sector they’re backing, Yang added, so they can assess technical feasibility, manufacturing constraints and customer demand long before the business has proven itself commercially. 

Not everyone is convinced that deep tech will provide a safe haven for investors looking for certainty amidst the AI chaos. Allen Lau, co-founder of Toronto-based deep-tech venture firm Two Small Fish, laughed at the idea of deep tech as a safe haven from the software downturn. “These sectors are hard to build,” he said. Plenty of software investors, he said, are inquiring about deals in the area, which includes robotics, quantum and medical devices, but few are piling in. 

The Canadian Robotics Council (CRC), a Toronto-headquartered non-profit that advocates for the robotics industry, is trying to make it easier to invest in a slice of the deep-tech market. In April, the council launched a capital committee to help venture firms better understand how to evaluate hardware-heavy businesses, with the goal to channel more money into robotics startups.

Ryan Gariepy, co-founder of Clearpath Robotics and CRC chair, said the group is hoping to capitalize on the growing interest in robotics and physical AI, which, he said, didn’t exist when he was building Clearpath in 2009. 

Despite the innate challenge of building and funding deep-tech companies, Yang said it’s an area Canada should lean into. The country has a strong record of producing top-tier scientific research and talent through its universities, she added, giving it an edge in the sector. “We never had any strategic advantage in software,” she said. “But deep tech, we actually have this asset.”