Jonathan Cohen, Founder & CIO of Robocap on physical AI, the next technological frontier

Jonathan Cohen shares his journey from building drones as a hobbyist to becoming one of the earliest pure-play investors in robotics and AI. He discusses how Robocap identifies authentic AI exposure, the critical role of technical due diligence, and why the convergence of AI and robotics, what he calls Physical AI, will redefine global industries in the decade ahead.

Jonathan Cohen has over 20 years of investment experience in thematic equities and over 10 years in robotics and AI. Following his graduation in Management with a Major in Finance from the University of St Gallen in 2004, Jonathan joined Bedrock SA, a multi-family office based in Geneva and London, where he was hired as the equity analyst for an activist investor. He then joined Goldman Sachs where he was promoted to Associate Investment Professional within the Institutional Wealth Management team. In 2009, he became the CIO of Huet & Cie BV, an Independent Asset Manager based in Geneva and Zurich, where he led a team of three analysts. Following Huet & Cie he took a role in London as Senior Portfolio Manager of the US team of London & Capital Asset Management where he led a team of three portfolio managers. He founded Robocap in 2015.

Back in 2016, Robocap launched a Robotics & AI-focused fund, well before the technology became a mainstream investment theme. What gave you the conviction to launch so early, and how has your original thesis evolved over the past decade?

I have always been fascinated by new technologies. Back in 2014, I was building drones and learning how to programme robots through online MOOC courses — a passion that gave me strong conviction in our central investment thesis: that we are entering a robotics and AI revolution which will redefine every industry.

As an investment professional, I began mapping out the investable universe and found many “under-the-radar” companies with solid fundamentals. I was excited to be among the first fund managers in this space — and the first with a pure-play approach.

Our original thesis remains intact, but some sub-themes such as AI have accelerated beyond expectations, while others, like 3D printing, have progressed more slowly.

You emphasize investing in “pure AI players”.. How do you differentiate genuine AI businesses from companies simply rebranding as “AI-driven” without meaningful revenue exposure?

We believe that investing in genuine AI businesses is critical to staying true to the theme. For example, investing in a company that uses only a simple algorithm — without true AI integration — would be misleading for investors seeking exposure to this disruptive technology.

We assess the share of company revenues expected to be driven by AI within the next two years, targeting at least 40%, while our portfolio average sits around 80–85%. Distinguishing authentic AI exposure from “AI-washing” requires deep technical expertise and continuous monitoring.

Robocap’s advisory board includes leading AI developers and academics. Can you explain how this collaboration influences your investment decisions, and perhaps share an example where their insight materially impacted portfolio allocation?

Our Advisory Board, which is made up of leading experts in AI & robotics, plays an essential role in helping us identify trends, assess emerging technologies, and conduct rigorous due diligence.

For instance, back in 2016, the board highlighted key innovations in AI and encouraged us to increase exposure to the theme — leading to our investment in Nvidia in early 2017. In another case, their insight helped us avoid investing in a company that falsely claimed to be active in AI. Their input ensures our strategy remains both technically and thematically disciplined.

Cybersecurity is becoming one of the fastest-growing applications of AI. What types of AI cybersecurity models or companies do you see as most investable, and how do you evaluate their durability against rapidly evolving threats?

Cybersecurity is fundamentally a data problem — and therefore an ideal field for AI.
AI is indispensable because:

• Detecting new threats requires pattern recognition beyond human speed.
• Response times must be instantaneous.
• The talent shortage in cybersecurity is structural.
• Attack surfaces continue to expand through the cloud, IoT, and AI systems themselves.

We see strong investment potential in AI-driven security platforms such as CrowdStrike or Palo Alto Networks, which integrate multiple modules, minimize gaps in protection, and help rationalize costs. Their competitive advantage lies in their proprietary datasets and the quality of their R&D teams.

In a world of shifting trade policies, regulation, and rate cycles, how do you assess macro risks versus the secular growth opportunity in AI?

Our conviction in long-term secular growth is the foundation of our investment process, and we seek investors who share that horizon.

Among macro factors, the rate cycle has been the most impactful — influencing valuation multiples and often difficult to hedge. Trade policy and regulatory shifts can be unpredictable, but diversification across industries and geographies helps mitigate those risks. Ultimately, macro variables can distort timing, but they do not alter the direction of technological transformation.

From healthcare AI, to logistics automation and space robotics, AI applications are broadening. Which verticals do you believe are at the tipping point of value creation for investors today?

AI is, in our view, the most important technology of our lifetime. Its impact now extends from software to the physical world — through AI agents, logistics automation, autonomous vehicles, and soon, humanoid robotics.

We call this new era Physical AI, where intelligence meets embodiment. This convergence will create unprecedented value over the coming decade.

Since inception, the Robocap UCITS Fund has delivered over +250%. How has your portfolio construction evolved over time, and what lessons did you learn from challenging years like 2018 or 2022 when markets were more volatile?

When we launched the Robocap UCITS Fund in 2016, our exposure was primarily in industrial robotics and general automation. Today, our portfolio is much more focused on AI and Physical AI.
Each challenging year has offered key lessons.

• In 2018, during the U.S.–China trade war, we learned the importance of overweighting software names, which are less exposed to tariffs.
• In 2022, rising rates compressed multiples across the sector — in hindsight, hedging our equity exposure would have been prudent.

We continuously adapt our portfolio construction based on such experiences while maintaining thematic purity.

You’ve spoken about the “new generation of physical AI.” With developments such as humanoid robotics, do you believe robotics will soon reach mass-market adoption, and how do you position the fund to capture this?

We believe the technology for autonomous vehicles is ready to scale today and will transform transportation over the next 5–10 years.

Humanoid robots, while not yet mass-market ready, are advancing rapidly and could reach large-scale adoption within 4–5 years. When that happens, they are likely to become one of the world’s largest industries — merging robotics, AI, and hardware in ways comparable to nothing else before.

You often stress the importance of being “on time” when investing in AI. What signals or KPIs do you look for to determine when a company is genuinely ready to scale AI-driven revenues?

From a qualitative perspective, we listen to customers and observe whether AI genuinely improves their workflows — or merely serves as a marketing layer.

Quantitatively, we track metrics such as revenues generated from AI modules and the conversion rate from testing to adoption. When clients are willing to pay for AI features, it’s a clear signal that value creation is real and scalable.

If we look forward to 2035, how do you expect the AI investment landscape to evolve? Will AI remain a specialist theme, or do you expect it to become a core allocation across equity portfolios, similar to how tech or healthcare are treated today?

AI is already a core allocation — even if unintentionally. Nvidia alone represents 8% of the S&P 500 and 14% of the Nasdaq, while AI-driven productivity accounts for roughly half of US GDP growth in 2025. The technology is complex and requires expertise as explained above, and many allocators are still thinking in terms of geographies and sectors in a global world, while passively waiting that the AI revolution joins the indices.

Much like electricity during the Second Industrial Revolution, AI will underpin virtually every sector. It represents the greatest and fastest transformation humanity has ever experienced.