In our opinion, artificial intelligence (AI) is the most transformational, long-term investment theme of a generation. Its far-reaching social and economic impact could be comparable to that of the railways, the internal combustion engine, the telephone or television.
At the moment, we don’t think all companies (or governments) understand the opportunities and threats that AI creates but many are beginning to wake up to its impact. Six years ago we spent a lot of time explaining what ‘AI’ is – that certainly is not the case now. We believe there will be winners and losers from the widespread adoption of AI, with the winners enjoying faster rates of growth than their peers.
The social and economic impacts of AI are potentially huge. AI will undoubtedly remove a lot of dull, repetitive, rules-based work – this should free up humans to focus on value added areas, or to think more strategically or creatively. However, we cannot ignore that many jobs – in transport for example – simply involve moving things from A to B, either on roads or within warehouses. In the developed world, many of those jobs are likely to disappear in the coming years due to AI-enabled automation. The challenge for policymakers will be to make sure that the transition period is not too painful, and that workers are given the tools and ample opportunity to reskill. This is not a new challenge or even specific to AI – it is a problem that has existed ever since the industrial revolution held out the promise of automated or mechanised work.
One of the key things to remember about AI is that it will disrupt any job or profession that is rules-based – so for example accountancy is one of the professions that could be most affected by AI. However, creative industries are not immune either – AI can already create music that is indistinguishable from music written by humans, and, as we’ll see later, it can even create art.
Whilst there are potential challenges that will emerge as AI is rolled out across the economy, there are huge benefits too, and we believe these will outweigh the downsides. For example, if we look at healthcare, AI is already, on average, better at analysing mammograms than the average human radiographer. This is important because it takes a human years to be trained and become skilled in radiography and there is obviously a limit on how much analysis a human radiographer can do in a given day. AI can work 24/7, if required, and play a critical role in clearing treatment backlogs and providing better clinical outcomes at a lower cost.
There are revolutionary environmental applications too: for example, Google has an AI unit called DeepMind and its work on protein folding has opened the door to ‘eco enzymes’ that could be used to break down the plastic and other contaminants that pollute the world’s rivers and oceans. Nvidia, one of the companies we invest in, has created a complete digital replica of planet Earth using AI and it can model how different environmental scenarios – such as those related to climate change – could play out over time. That kind of technology could revolutionise how we think about climate change and how policymakers respond, and crucially it can give us early warning of specific climate-related disasters so that we can take action (e.g. evacuating a town or city or reinforcing flood defences) ahead of time. We mention this stock for illustrative purposes only, and it's not a recommendation to buy or sell yourself.
The role AI can play in a diversified portfolio
We think AI should play a role in all diversified portfolios – even if you don’t agree with or like the deployment of AI, it is here – the genie is out of the bottle. Five or ten years ago things like self-driving cars and ‘doctor bots’ sounded like something from a science fiction film – now they are a reality.
We’d urge investors to think about how AI will affect the earnings and long-term growth trajectories of the companies they invest in – companies who use and deploy AI can build even stronger moats around their existing businesses. Those that don’t engage run the risk of potentially becoming irrelevant. For example, in the US we’ve already seen some educational companies warn that AI has had a negative impact on their business, as the rise of generative AI (and particularly the emergence of ChatGPT) has led to reduced demand for traditional learning tools.
Latest developments in AI
The advent of ChatGPT – perhaps the first global watershed moment in AI – took some people by surprise but actually it is just part of a longer-term trend towards better and better Natural Language Processing (NLP). NLP forms the crucible within which those looking to assert their AI dominance have sought to burn most brightly.
To give some context here, ChatGPT was launched in 2018 with 150 million parameters (available connections to piece information together), and has increased in complexity to the point that GPT-3 now has 175 billion parameters – in other words, more than 1,100 times more parameters today than just four and a half years ago. ChatGPT uses “GPT3.5”, a tweaked version of GPT-3 with broadly similar capabilities. Microsoft (MSFT) paid OpenAI for exclusive access to commercialise GPT-3 in Q3 2020 – it’s one of the key reasons that we own a position in MSFT. We mention this stock for illustrative purposes only, and it's not a recommendation to buy or sell yourself.
ChatGPT’s capabilities are undoubtedly impressive, but there are other equally impressive AI platforms out there which haven’t created the global furore that ChatGPT has. For example, in the following link, Dall-E was asked to imagine the broader context for a number of famous paintings, and then “paint” the larger canvas in the style of the relevant artist. You can see the outcome for Vermeer here: https://openai.com/blog/dall-e-introducing-outpainting/
How to invest in AI
Stepping back a little, our core thesis is that we expect AI to permeate the entire economy over time. This is something that all investors should think about – not least because some businesses will face profound disruption because of AI.
To access the widest range of opportunities, investors should look for funds and strategies that invest on a global and unconstrained basis – not just for reasons of portfolio diversification but because different countries have different levels of adoption and engagement with AI. For example, in China AI has been part of everyday life for many years and it has been adopted at a level that would most likely prove difficult in the West due to ethical and privacy concerns.
It is also important to remember that AI is only as good as the data that it is trained on, so companies and institutions that own or control rich, deep and complex data sets can play a vitally important role in the ongoing development and deployment of AI – if of course they choose to engage with it.
Stocks mentioned for illustrative purposes only, and not a recommendation to buy or sell yourself. Forecasts are not guaranteed. Always seek financial advice before investing.