Recent research trips to Tokyo and San Francisco have reinforced just how central artificial intelligence (AI) has become to today’s investment debate.
On the US West Coast, the conversation is dominated by rapid advances in AI models and the race to build ever more powerful data centres.
In Japan, the tone is quieter but equally interesting, with discussions focused on engineering excellence, industrial strength and businesses that underpin the physical workings of the global economy.
For us, the question is not whether AI matters. It clearly does. The more important question is how to invest sensibly during a period of rapid technological change, while staying grounded in long-term value creation and cash generation.
The scale of investment
The world’s largest technology companies are currently investing heavily in AI-related infrastructure. Businesses such as Amazon, Alphabet (Google), Microsoft and Meta are collectively spending hundreds of billions of dollars each year on advanced semiconductors, large-scale data centres and the power systems required to operate them.
These facilities are vast in scale and represent a significant commitment of capital. Forecasts suggest that, over the coming years, cumulative investment in AI infrastructure could run into trillions of dollars.
As long-term investors, our role is to step back and consider what this might mean. AI is already generating meaningful revenues in certain areas, and adoption has been rapid in fields such as coding assistance and data analysis. At the same time, the revenues today are still relatively modest compared to the scale of capital being deployed.
AI may prove to be one of the most transformative technologies of our time but for investors it also puts into sharp focus the need for balancing enthusiasm with careful analysis.
What history teaches us
Technological revolutions rarely follow a simple, straight line. The railways, electricity and the internet all reshaped economies and improved living standards. Yet the journey for investors was often uneven.
The late 1990s offer a useful reference point. Large sums were invested in building out the infrastructure of the internet. Some companies eventually delivered extraordinary returns but many early participants struggled when expectations ran ahead of commercial reality. In many cases, the most successful business models emerged after the initial wave of infrastructure investment had taken place.
AI may follow a different path and it may develop more quickly. Nonetheless, history encourages a degree of humility. Innovation can create enormous value but that value does not always accrue immediately or evenly.
How we are positioned
We aim to participate in long-term structural trends without becoming overly reliant on a single theme. While we do not currently hold Nvidia, one of the most visible beneficiaries of the early AI infrastructure build-out, we do have meaningful exposure to the broader ecosystem through several high-quality businesses.
These include:
• TSMC (Taiwan Semiconductor Manufacturing Company) – a global leader in advanced chip manufacturing, serving customers across smartphones, computing and AI applications.
• Amphenol – a specialist in connectors and sensors used in data centres as well as a wide range of industrial markets.
• Schneider Electric – a leader in electrical distribution and power management, supplying essential equipment to data centres and infrastructure projects.
• Microsoft, Amazon and Alphabet – major cloud providers that are central to AI development, while also benefiting from diversified and well-established business models.
In each case, AI represents an additional growth driver rather than the sole reason for owning the company. These businesses have strong competitive positions, diversified revenue streams and proven ability to generate cash.
The ‘reassuringly physical’
Alongside these AI-related holdings, we continue to invest in businesses that are firmly rooted in the tangible economy. We sometimes describe these as ‘reassuringly physical’ companies whose products and services meet everyday essential needs.
One example is MonotaRo in Japan, an online distributor of maintenance, repair and overhaul (MRO) parts. It supplies millions of items to businesses across the country, helping factories and workshops keep machinery running efficiently.
While the company benefits from digitalisation trends, its core purpose is straightforward: machines need maintenance and parts need replacing, regardless of the latest technological developments.
In the United States, Federal Signal manufactures specialist industrial vehicles such as street sweepers and sewer-cleaning trucks. These products serve municipalities and utilities where demand is driven by essential infrastructure maintenance.
Roads must be cleaned and water systems maintained whatever the economic backdrop – and however sophisticated AI becomes.
These kinds of investments reflect an important aspect of our philosophy. While we recognise the potential of new technologies, we also value businesses operating in well-defined niches with steady demand and durable market positions.
Maintaining balance
AI is likely to influence many industries over time and we expect new opportunities to emerge as applications broaden and commercial models evolve. At the same time, periods of rapid innovation can also bring volatility and elevated expectations.
Our objective in managing Brunner is not to predict precisely how every technological development will unfold. Instead, we focus on building a diversified portfolio of high-quality companies – some participating directly in technological change, others grounded in essential physical activity – that together can deliver sustainable, long-term returns.
In times of excitement, maintaining that balance is more important than ever.
Julian Bishop is co-portfolio manager of the Brunner Investment Trust. The views expressed above should not be taken as investment advice.