AI, deep tech too new for private equity, says family office investor

Pre-2015 deep tech is inferior to advances made in recent years, which makes venture capital the preferred mode of investing for Ferretto Capital
AI, deep tech too new for private equity, says family office investor

Deep tech applications only became viable seven years ago, making it more suitable for venture capital (VC) investments compared with private equity (PE), according to a senior executive at Ferretto Capital, who added that within VC, the family office is eyeing industries such as healthcare, logistics and food services.

James Wang,
Ferretto Capital

“Around 2015 or 2016 was when many AI (artificial intelligence) benchmarks finally reached human-level capability on specific tasks, especially computer vision. That was when genetic sequencing costs fell to a point where a lot of synthetic biology applications became viable,” said James Wang, chief portfolio strategist of Ferretto Capital, a single-family office helmed by chief investment officer Champ Suthipongchai.

“Basically, anything created before that time is fundamentally different or inferior to what came after. PE is now mature – and interesting – for software startups, but what is now classified as ‘deep tech’ is too new for PE to usefully capitalise until the market and application is obvious and much of the returns have already been realised,” he told AsianInvestor.

Deep tech refers to the development of engineering or scientific innovations for technology that underpins a variety of applications such as clean tech, energy efficiency, smart homes, semiconductors, computing architecture and healthcare.

Wang spoke recently at the AsianInvestor Thematic Investing Forum 2022 about investing in a digitised and automated future.

Automation and robotics will help with the labour shortage that several economies face due to issues such as an ageing population, he said during the event. And, as with other deep tech advancements, it has only been relatively recent that decent developments have been made.

“It's only been in the past a year or two, that we've actually had a lot of the developments within artificial intelligence in terms of reinforcement learning,” he said.

“With some of these other technologies that have just gotten here, you can actually have viable robotics companies, that can actually replicate a lot of the things within again, construction, food services, all of these real sectors, and actually be able to go and do these tasks.”


“If one simply listens to startup pitches from the ‘bottom up’, then one is just investing in charisma and not actual technological and business potential.”

When investing in an area as complicated and technical as deep tech, he takes a macro- and market-first approach to understand which technologies are commercially viable and have engineering risk as opposed to research and development (R&D) risk.

Some examples he gave in a separate interview with AsianInvestor were quantum computing, AI and gene therapy.

“There have been multiple waves of these technologies all the way back to the '80s and '90s. The problem then was that the potential looked great, but they were just ‘one or two breakthroughs’ away from commercial viability. We think trying to ‘front run’ technological progress is a losing battle. You take R&D risk that is uncontrollable, unpredictable, and creates risk of complete losses when the technology never reaches commercial viability,” he said.

Instead of investing in startups that focus on technologies with R&D risk, he prefers those that have engineering risk, which provides a “finite, estimable development” time.

To gain an understanding of these risks and the industry’s macro trends, he recommends that investors rely on partners, experts, advisors and managers who can provide the technical expertise and recognise the difference in engineering and R&D risks.

“Once those things are known, then an investor can look at companies within these different spaces,” he said. “If one simply listens to startup pitches from the ‘bottom up’, then one is just investing in charisma and not actual technological and business potential.”


Healthcare and robotics are areas that Wang is most partial to.

“Healthcare has huge potential, though one must be sensitive to the regulatory requirements within it, which can be quite extensive,” he said.

“We most like logistics, food services, cleaning, construction, and similar manual labor tasks that have a massive secular tailwind,” he added. “There simply are fewer young, strong bodied humans to throw at these problems relative to older humans at this point. Given that, we either need to have fewer of these tasks done (lower standard of living) or massively increase productivity with things like robotics.

Wang and Suthipongchai are also general partners of a San Francisco-based venture capital firm called Creative Ventures.

They have several deep tech companies in their portfolio including quantum computing firm Bleximo, AI construction scheduling firm Alice, 3D scanning analysis firm Riven and Imvaria, which uses AI to drive digital biopsies for lung diseases.

"The distinction with these particular areas is that there is enough variability — each hotel room that needs to be cleaned, or pizza to be prepared is slightly different — that traditional automation methods without any AI could not perform them," he said.

"Now that AI has reached a stage that these variable, but still largely repetitive tasks, can be performed, these areas suddenly have massive potential to be economically transformative—and incredible financial opportunities."

Wang declined to share the family office's risk and return targets only that they are long-term. "The best lesson we can share is to have a good strategic allocation base, but even more importantly, to have high, uncorrelated 'alpha' managers who can provide outstanding returns in any environment."

"We like early-stage VC for this reason, since disciplined managers are more insulated from short-term variation in public markets. Additionally, the early-stage tends to be more inefficient and can provide better risk-adjusted returns relative to the late-stage—which, at this point, has basically became what the public market used to be, given how many companies are staying private but continuously raising massive rounds," he said.

¬ Haymarket Media Limited. All rights reserved.