By Chris Iggo, CIO core investments, and Jeremy Gleeson, portfolio manager at Axa Investment Managers
Artificial intelligence has been a hard topic to miss in 2023. During Q1 earnings calls, 110 constituents of the S&P 500 index mentioned ‘AI’ – the highest number in more than a decade.
But AI is pervasive and has been for some time. Algorithms dominate our lives, from music to video streaming, and from online shopping recommendations to much, much more.
There are considerable investment opportunities in both the design and infrastructure of AI, but also across a broad range of sectors as companies begin to use this technology to improve their own business models.
See also: Central banks catch their breath: Will biotech shine again as rates stop rising?
In order to effectively engage with these opportunities, investors must have at least a top-line understanding of the various nuances in AI design and application. So let’s begin with a brief explainer.
At its most basic, AI is simply ‘machine learning’ – a computer’s ability to think, learn and perform functions we associate with humans, establishing rules or algorithms by which to operate. AI can also ‘reason’ – by deciding on algorithms – to choose particular outcomes and information.
AI comes in several subsets. Familiar household virtual assistants such as Amazon’s Alexa or Apple’s Siri are subsets of machine learning, where a neural network attempts to simulate human knowledge acquisition. This is done by creating algorithms trained to learn from experience.
Deep learning is another subset of machine learning which involves training neural networks to mimic the human brain structure. Neural networks typically have multiple layers and can be trained to perform specific tasks such as image or speech recognition. Driverless cars and facial recognition are good examples of this.
See also: Digital infrastructure trusts: Tapping into the fourth industrial revolution
Generative AI, such as ChatGPT, takes all this much further. It can create and produce text, images, video, and numerous other types of content. This could disrupt a vast range of industries.
Growth potential
While there is a lot of buzz around AI right now, it is very much a long-term productivity story. It will be a vital investment theme over the coming decade and beyond, but its impact will be incremental.
Right now, cloud computing, data centres, server and computing capacity, as well as semiconductors and semiconductor equipment companies, are among the potential winners of generative AI.
Chipmaker Nvidia became a member of the select club of trillion-dollar companies when demand for its graphic processors which train AI systems rocketed this year. Another chip manufacturer, Marvell Technology, has also seen strong share price gains after an upbeat earnings report, driven by the excitement engulfing the AI sector.
Banking, high tech, and life sciences are among the industries that could see the biggest impact as a percentage of their revenues from generative AI, but many others will feel its force.
In addition, the current focus on infrastructure will expand over time into companies creating applications for this technology – and the possibilities here are arguably endless, from diagnosing diseases and improving energy efficiency, to countless other uses we have yet to envisage.
Looking at the companies exploiting this technology, software-based client relationship management firms such as Five9 and Salesforce are two examples already adopting AI, while services companies such as Accenture, Capgemini and Globant are involved in the development of applications around AI, spearheaded by the larger players including Amazon, Microsoft and Alphabet.
And it’s not just tech firms wanting to get in on the action. Since its theoretical birth in the 1950s, more than 340,000 AI-related invention applications have been filed. But over the past decade, the rate of applications has rapidly accelerated, from 2,560 patents in 2010 to over 140,000 in 2021.
Cautious approach?
There should still be some caution among investors, of course, as not every AI innovator will be a success, and certainly the hype around AI will bring back memories of the dotcom bust.
That being said, today’s technology sector is a different beast. Back then, in several cases, company valuations were based on aspirations rather than anything fundamentally tangible. Today, companies have real clients, revenues and profits. Many are reaping the benefits of the current wave but it is too early to assess what the environment will look like in five, let alone 20, years’ time.
There will be issues to be resolved and scrutinised, as there are for all new areas of technology. Security, accuracy and consistency are chief among them. And regulation will be a key factor here. The EU is already developing an AI Act in a bid to protect companies and consumers.
See also: Investors must be wary of ‘AI losers’ as tech valuations spike
Indeed, the velocity to reach commercialisation will likely be swift, especially as consumers at home are already adapted to conversational AI user interfaces such as Alexa, Google Assistant or Siri. This should help speed up adoption and underlines the structural changes already underway, bolstering AI’s long-term investment potential.
For investors, particularly those with a bias towards growth equities, AI offers tremendous potential opportunities to benefit from superior growth in earnings. At the same time, some companies will lose market share if they fail to exploit opportunities offered by more powerful computing.
In developed countries facing demographic challenges and labour market constraints, the power to automate economic activities and deliver services and goods to consumers using AI techniques will have revolutionary economic and social implications. We believe this will be a dominant investment theme for years to come.
