Artificial intelligence (AI) has already been identified as the fourth industrial revolution, a concept first recognised by the World Economic Forum seven or eight years ago, noted Alexandra Spasov, thematic funds analyst at Quintet Private Bank. This was even before ChatGPT brought AI to public awareness.
Like previous industrial revolutions, AI is expected to profoundly transform daily life, businesses, production and consumption. It has an infinite array of possibilities and adaptations applicable to all economic sectors. Spasov observed that its adoption rate has been rapid, increasing from 5-6% at the end of last year to 14% at the end of the current quarter.
The thematic funds analyst identified several key investing phases in the AI ecosystem involving:
Infrastructure: initial investments focused on foundational infrastructure, beginning with semiconductors, where companies like Nvidia are leaders. The next step is cloud infrastructure, dominated by providers such as Amazon, Microsoft and Google, which host the vast amounts of data AI relies upon. For Spasov, the sector has already borne fruit.
Energy: AI operations--exemplified by ChatGPT, which requires 10 times more energy per query--“demand monstrous energy consumption.” This makes investing in energy infrastructure improvement and optimisation a crucial aspect of leveraging AI growth. This also ties into national security, as countries aim to have data centres and energy sources on their own territory to protect sensitive data. France, England and Germany are actively pursuing this, focusing on smart grids and energy storage solutions like batteries.
Applications/beneficiaries: after infrastructure, the focus shifts to companies directly benefiting from AI implementation, known as "champions,” such as Amazon in e-commerce, some Chinese auto manufacturers active in the autonomous driving technology and companies amassing large amounts of data in the healthcare sector.
AI applications are seen as infinite, spreading to every part of the economy. But companies that fall behind may face greater challenges, much like the struggles Kodak experienced when it failed to embrace digitalisation.
Active or passive investment in AI
For highly concentrated sectors like semiconductors, where the investment universe is small, Spasov considers passive solutions such as exchange-traded funds (ETFs) to be a suitable option.
Thematic investing is a long-term investment… over five to 10 years where the allocation is strategic and partially tactical
However, she often prefers active managers for the broader and rapidly evolving AI theme. An active--human--asset manager (some funds select securities by using AI) offers more flexibility and agility to adapt to rapid technological evolution and regulatory changes. They can strategically shift investments from early-stage infrastructure to later-stage beneficiaries for optimal performance.
Spasov observed that some thematic ETFs struggle to adapt quickly and may lack “pure play” of the thematic investment. They are often, in fact, “global tech” funds, as they overweight large “Magnificent Seven” companies while missing smaller differentiating firms that may generate outperformance. “Thematic investing is a long-term investment… over five to 10 years where the allocation is strategic and partially tactical.”
Fund selection approach
Quintet prioritises funds offering diversification across industries like healthcare, robotics and other productivity-enhancing sectors, looking through the whole value chain with a focus on the most promising sub-segments. The bank proposes a comprehensive global solution in a single fund or seeks funds that offer dedicated solutions for sub-themes (like infrastructure or cloud).
A significant criterion is the fund’s historical presence, preferring those that have existed for five to six years, indicating an early understanding of AI rather than just joining a trend.
Ensuring the purity of AI investment
To distinguish true innovation versus hype companies that are merely marketing themselves as “AI-driven,” Spasov remarked that fund managers verify a company’s genuine involvement in AI by setting thresholds or quantitative filters for revenue generation directly tied to AI (e.g., 50% of revenue).
They also use qualitative checks to confirm revenues originate from AI, preventing misrepresentation. She stressed that in the semiconductor industry, only a subset of listed companies is truly linked to the AI market. “The universe of relevant candidates counts only 40 to 50 stocks for active portfolios.”
Must-haves
Questioned on the essential AI firms that frequently appear in AI-focused investment funds, Spasov observed that this list often includes Alphabet (Google) and Microsoft for their broad AI leadership. Other key players are Fortinet (a "pure player" in cybersecurity), Nvidia and TSMC (semiconductors), Equinix (data centres) and Fanuc (robotics and automation), along with Precision Medicines in robotic surgery.
Key sub-sectors potentially underestimated
Spasov commented that the cybersecurity sector is “consistently underestimated” because it is always playing catchup against increasingly sophisticated AI-driven attacks. Investments in cybersecurity budgets need to grow significantly, at least matching AI investments, to adequately protect companies.
Moreover, given her frequent references to the sector during the interview, it is not surprising for Spasov to flag the subsegment--humanoids and mobile robots--as undervalued. Beyond industrial applications, these robots, equipped with AI-powered cameras, are envisioned as “living data collectors.”
She expects that they will gather and process vast amounts of real-world data, “transforming robotics companies into data-driven businesses that resell this valuable information.” This concept also extends to autonomous vehicles as mobile data-gathering platforms.



