AI: A giant leap for mankind

This article was originally published in Portfolio Adviser.

Some information contained herein has been obtained from third party sources and has not been independently verified by Polar Capital. Neither Polar Capital nor any other party involved makes any express or implied warranties or representations.

Xuesong Zhao and Nick Dumas-Williams of the Polar Capital Artificial Intelligence Strategy explain how AI is changing the world and explain that capturing the opportunities will require a much broader investment focus than the immediate technology market.

How differentiated are artificial intelligence (AI) strategies from traditional technology strategies?

Nick Dumas-Williams (NDW): When we launched this strategy back in 2017, our belief was that artificial intelligence (AI) would change the world. Since the launch of ChatGPT, and the many other major developments since, we believe that AI is now the next general purpose technology.

Such advancements are both very rare and incredibly impactful, and they ultimately change how every single industry operates. This is the core tenet of why the strategy was established, recognising that opportunities will emerge across all sectors, and that capturing them requires a broader investment focus than the immediate technology market.

Xuesong Zhao (XZ): Within this strategy we invest in both the enablers and beneficiaries of AI. The enablers are currently more aligned with the technology universe and represent less than half of the strategy, while the beneficiaries are largely non-technology companies that we expect to win through AI adoption.

We have seen in the past that sometimes the greatest opportunity to benefit from a technology is investment in these ‘downstream’ beneficiaries. It is an imperfect parallel, but one of the best ways to play the emergence of wifi was to invest in Starbucks. Starbucks used wifi to fundamentally transform the coffee shop, using free internet as a customer acquisition tool and larger shops to capture more traffic.

Our aim with this strategy is to capture these Starbucks-type transformations in markets that we believe could offer far larger long-term opportunities than the tech angle alone.

Is the success of NVIDIA overshadowing other opportunities in the AI theme?

NDW: Right now we are in the midst of an accelerated AI infrastructure build-out which has been driving exceptional earnings growth for NVIDIA. However, while NVIDIA is a central pillar of the AI compute stack, it’s not the only opportunity within AI because other parts of the computing stack are being completely overhauled at the same time.

Fundamentally, existing computing architecture is poorly suited to AI workloads. We consider traditional computing infrastructure as akin to a family car, and the new AI computing architecture as Formula 1; not only is the engine completely different, so too are the rest of the parts in the system.

XZ: Most AI work so far has been focused on the initial phase, the training of AI models. This is particularly suited to parallel computing in which NVIDIA dominates. As such the majority of this early growth has come to NVIDIA, and we would not expect to see those dynamics to change significantly in the near term.

However, we see plenty of opportunities beyond NVIDIA due to the wider infrastructure changes taking place, especially as AI workloads move beyond the training phase to business and consumer usage, those workloads known as inference.

NDW: Away from infrastructure, we are starting to see greater recognition of AI winners within the beneficiaries. The first AI products and case studies of adoption are emerging with very promising results.

These are already delivering material earnings power to AI leaders within different industries, and this is slowly being recognised by investors. We have a significant portion of the portfolio invested in companies that we believe can accelerate their earnings growth through AI and are excited for this to become more widely understood and reflected in stocks in the near future.

What are some of the ways you expect AI to disrupt non-technology sectors?

NDW: The focus so far has very much been on augmentation; namely how can AI help us as humans do our jobs better? AI captured public attention initially in creative tasks, but we think the real opportunity is increasing productivity, reducing the number of menial tasks we do and the errors we make.

The further we go into replacing manual, repetitive work, and freeing up people to do higher value work, the greater the returns potential for AI and it will apply to all sectors of the market. The early examples we are seeing are remarkable and suggest that AI can match, or even surpass, the hopes that many have for delivering these productivity gains.

XZ: AI is also starting to disrupt areas people wouldn’t necessarily consider, impacting wider societal infrastructure. For example, we are seeing interesting applications of AI to solve problems like power intermittency with renewable energy, balancing the grid to deal with different conditions at different times of day.

At a time when questions are being asked of the power requirements of AI, we think it is underappreciated as part of the solution.

Which sectors are you currently favouring when seeking AI beneficiaries?

NDW: Given our excitement in the potential for AI to transform all sectors, it is almost easier to speak to areas to which we currently under index. We have seen remarkable progress in a short space of time, with companies across all sectors embracing the potential to transform their businesses through AI.

This is reflected by the portfolio having holdings across most sectors. We are invested in AI leaders that are demonstrating tangible growth accelerations and profits across industrials, financials, advertising, legal and healthcare, where their adoption of AI is demonstrating the ability to generate differentiated returns.

Two areas where we have high longer-term hopes but moderated near-term expectations would be banks and drug discovery within healthcare. Banks are almost the poster child for old infrastructure, with legacy technology stacks that in many cases are complex to overhaul and not conducive to digital transformation.

While AI has a huge opportunity to solve some of these problems, the rates of adoption and the ability for AI to move the needle in the near future are also more limited because of these reasons. Hence it is a sector we are keeping a close eye on, but there are better opportunities elsewhere.

Drug discovery in healthcare has long been hoped to benefit from AI advancement given the scientific and financial challenges in bringing new drugs to market. AI-led drug discovery certainly has huge promise, however from an investment perspective we are more cautious because so many of the assets are early stage, with the associated potential initial high failure rates as both the industry and the technology mature. We are encouraged by continued progress and await further clinical data with interest.

That said, AI is helping to solve many smaller bottlenecks within healthcare, improving standards of care such as the accuracy of imaging, relieving administrative burden on doctors and easing workloads. With global healthcare systems under such strain, we are excited about the potential for perhaps more incremental improvements to come to something meaningful.

Is there a company investors might not traditionally associate with AI, but which you are investing in?

XZ: One example which has been starting to garner more attention is a company we hold called Axon Enterprise. Axon Enterprise makes body cameras for police and public services and have recently released an AI product called Draft One.

By some estimates police officers spend up to 40% of their time writing up incident reports. Draft One can automatically generate a police report from the video and audio captured by the body cam. It has been estimated this could reduce the time taken to write reports by over 80%, a productivity improvement that helps to significantly reduce staff turnover and burnout.

This captures why the AI opportunity is ultimately so interesting, it is not just the incremental profit on offer, but the ability to solve for a different problem that technology has not been able to address before.

Why is now a good time to invest in the AI theme; are valuations still supportive?

XZ: Some investors believe AI is in a hype cycle, we would strongly disagree. AI was first developed in the 1950s, with the first applications used for weather forecasting. We believe that the hype occurred in the following 60 years, and the 2017 advent of the transformer model, a new class of AI models that enabled ChatGPT, represents the inflection point.

The capabilities of the technology since that point continue to grow exponentially. Having hit this inflection point, the rapidly accelerating rates of broader adoption we are now seeing typically signal a sweet spot for investors.

NDW: It’s twofold really. The first is from an infrastructure point of view, which is still a significant portion of the portfolio. It’s our belief that we are at the beginning of the accelerated computing infrastructure build-out for AI. The reports we’ve seen show us this isn’t slowing down; capex is being revised higher and the industry growth remain very strong.

The next generation of NVIDIA chips are emerging, and the smartest people in the industry are telling us this infrastructure will unlock things we haven’t yet been able to do. We think the data suggests that this is the next general-purpose technology, similar to steel, electricity or the internet, and that we are on track for a monumental step forward for society.

The second is the opportunity for the AI beneficiaries. We are still in the first generation of many generative AI tools and soon many people, in some form, will have AI contributing to parts of their lives. These first tools are delivering real utility and productivity gains, so are already delivering real, commensurate earnings growth for companies too. We see a very rare opportunity to invest alongside a technology with the potential to upend all industries, and are excited about the investment opportunity on both a near-term and multi-year view.