Baillie Gifford's formula for investing in the age of AI 

Kyle McEnery has been thinking about AI longer than most. The Baillie Gifford investment manager traces his journey back to 2015, when he noticed his peers with physics PhDs pivoting en masse to machine learning.

By 2016, he had established an internal AI research team at the firm and led it for six years. Then, in late 2022, everything changed.

“GPT-3 came out, which I think was the world’s first introduction to generative AI and these large language models,” McEnery says. “And for me, this was a huge moment. I was used to using machine learning or AI, but this was totally different.”

Since then, he has spent time with the founders and executives of Anthropic and OpenAI, deepening his conviction that what is unfolding is not a cycle but a structural multi-decade shift.

For advisers trying to position clients in this space, the question is not whether AI is real. It is whether portfolios are built to capture the full duration of what is still, by McEnery’s reckoning, an early-stage story.

The market keeps getting surprised, and it should not

McEnery’s view is that the direction is broadly right, but the process is reactive and inconsistent.

“I think we’re absorbing AI’s impact in a very volatile and almost discrete way, reactive. We’re consistently underestimating the demand here. We get surprised on the supply side. We’re getting surprised in terms of when new applications come online.”

For advisers fielding client questions about AI-exposed positions, the volatility is not necessarily a signal of structural weakness. In many cases, it reflects a market that keeps underpricing what is coming.

The SaaS selloff was a first-order response to a genuine disruption signal. But McEnery’s observation is that the second and third-order reading is far more nuanced: many of the companies the market sold down indiscriminately actually stand to benefit from AI rather than lose to it.

His portfolio holdings in Samsara and Axon illustrate the point. Both are hardware-led businesses with software layered on top. The competitive moat is physical, built over years of deployment. AI cannot replicate that overnight.

“They’re excited about AI,” McEnery says of the founders and management teams he spends time with. “They’re not thinking, oh, AI is making my business go away. They’re going, oh, thank God AI has come. Because there’s so many things that I couldn’t do that now I can do.”

What to look for

McEnery runs a high-conviction, concentrated global growth portfolio at Baillie Gifford. The strategy is not trying to own every layer of the AI stack. It is trying to own a handful of exceptional companies with extremely long durations of high growth.

“The idea is that we’re looking for companies that have extremely long duration of high growth,” he says. “We’re looking for the most extreme examples of this.”

The reference point he keeps returning to is Amazon, a long-time holding in Baillie Gifford’s Long Term Global Growth strategy. Not because of what Amazon did in e-commerce, but because of its habit of stacking successive growth S-curves on top of each other, e-commerce, then cloud, then AWS, and now AI.

“We’re looking for companies that we think hold extremely strong competitive advantages and that we think will continue to adapt and open up new growth opportunities,” he says.

Hyperscalers and disruptors are not mutually exclusive

On the question of whether established hyperscalers or new disruptors will capture AI’s value, McEnery does not see it as a binary choice.

Hyperscalers hold advantages in compute, balance sheet scale and distribution that are difficult to replicate. But frontier models like Anthropic are becoming platforms in their own right, and McEnery sees a symbiotic relationship developing rather than a winner-takes-all dynamic.

Emotional discipline

The investment industry is filled with smart people who find it easier to be right by being negative than by being optimistic about uncertain outcomes.

“Volatility is a feature, not a problem,” he says. “Amazon, Nvidia, these companies which have generated over 100x return over holding periods, they have multiple drawdowns. And that is difficult as an investor.”

McEnery’s response to that difficulty is pragmatic optimism. Review the thesis, stress-test the competitive advantage, but stay open to uncertainty in a positive rather than a defensive way. “We open ourselves up to the uncertainty, but in a positive way. And that’s where we can actually get the best rewards.”

Returns in growth investing rarely go up in a straight line, but for those able to think long-term, McEnery believes the outsized returns his approach has delivered speak for themselves.

For advisers helping clients manage that same tension, McEnery’s approach is less a portfolio strategy and more a mindset worth adopting.