How generative AI has re-ignited investor enthusiasm for tech stocks.
John Gladwyn, Senior Investment Manager Pictet Asset Management.
While the ideas behind AI have been around longer than most investors have been alive, an area which had previously been regarded as the preserve of computer science departments has gone mainstream.
But even if tech stocks have experienced a strong rally, the AI revolution is still in its very early stages.
The key foundation models – the neural networks of AI trained on huge data sets – are rapidly increasing in size, delivering major gains in capability with each iteration. It will be interesting to see to what extent Google’s new Gemini model, which is due to be released this autumn, is an improvement on the current large language models (LLMs).
The wave of interest in generative AI is testament to the core attractions of the technology sector and its ability to find new growth.
In many ways, the burst of innovation taking place today has parallels with the rise of the Internet in the 1990s. The Internet was relatively slow to take off as it took time to build up connectivity around the globe; in 1990 only half a percent of the global population was online. Today, most people have a smartphone and an internet connection. ChatGPT had more users after two months than there were internet users in 1996.
Where are the opportunities?
The generative AI wave has only just begun; it will create opportunities across the digital investment universe in the years to come. Microsoft’s Bill Gates captured the importance of LLMs: “I knew I had just seen the most important advance in technology since the graphical user interface.”
Just like with the rise of the Internet, at present most of graphic the money being invested into AI is going into infrastructure. Companies like Google, Microsoft, Amazon, Meta and Tesla are still constrained by the availability of high-performance graphics processing units (GPUs) that are needed for both model training and inference. We believe that over time investment will broaden out beyond pure infrastructure, as the intelligence of these AI models is rolled out into applications.
The potential breadth of these applications looks almost limitless today. Healthcare drug discovery and diagnosis, education, art, finance – all will be transformed by AI in time. And that in turn will fuel demand for other tech sectors such as software, hardware and semiconductors.
Microsoft’s GitHub Copilot (for code generation) is one of the most successful scaled use cases for AI today. Global companies report a 20-40 per cent increase in developer efficiency from using Copilot, at a cost of around USD230 a year for enterprise users. Soon we will have Copilots for everything.
One thing is different this investment cycle – in many cases we think that this technology cycle favours incumbents over new entrants. Today, established tech companies are frequently the ones leading in AI. AI requires large amounts of data and training AI models is extremely expensive; both of which are naturally easier for large, scaled companies than for start-ups. Similarly, anyone can integrate with the LLMs – there is no natural advantage to startups here. Finally, AI favours companies with large existing user bases, since new AI product capabilities will be easier to roll out across well-established products with large user bases. For investors, that means there are many attractive opportunities to gain exposure to the AI theme through well-invested, well-run listed technology companies.
Bumps in the road
While the long-term promise is there, tech consultancy Gartner recently summarised the feeling of a number of commentators (and investors) when they said that generative AI was currently at the “peak of inflated expectations”.
As time goes by, we believe investors will become more selective. It will become increasingly important to understand the nature of individual products and the positioning of the different companies. We also think that time to revenue is an important issue. This is the kind of work we are focusing on as a team today, sorting through companies where the investment returns are justified by the hype.
We are clearly at the beginning of another major technology shift, one that will transform most technology (and non-technology) markets over time. It is inspiring to see how quickly companies are adapting to take advantage of AI. We look forward to seeing the break-out products that start to define this new AI era.
