A Conversation with Tom Henriksson: Exits, AI and Quantum After 2025
In November 2023, Oleksandr Komarevych, Editor of Startupreporter.eu, sat down with Tom Henriksson, General Partner at OpenOcean, to discuss the challenges and opportunities awaiting venture capital in 2024. One year later, in November 2024, they reconvened at Slush to reflect on how those predictions unfolded and to look ahead to 2025.
In November 2025, Oleksandr and Tom met again in Helsinki — this time to revisit Tom’s predictions for 2025, made a year earlier, and to explore how reality compared with expectations. They discussed the return of IPOs, the surprisingly hot market for quantum companies, the ongoing difficulty of raising new funds, and how OpenOcean is adapting its strategy by going earlier into AI-driven opportunities.
The conversation also looked beyond AI to quantum computing as a possible “next wave”, and touched on OpenOcean’s internal evolution, with a new generation of partners taking the lead on many of the firm’s latest deals.
Exits in 2025 vs. Predictions
Oleksandr: – In our November 2024 conversation, you anticipated more exits in 2025. Now that we’re at the end of the year, how did that unfold? What surprised you most about how 2025 actually played out versus your predictions?
Tom: – In broad terms, it went more or less as I expected. Bigger exits started to happen again. IPOs came back in 2025. We’re not at 2021 levels of volume, but there are clearly more IPOs now, and it looks like that will continue into 2026 at an accelerating pace, which is good.
For us at OpenOcean, we had a couple of smaller strategic exits. Nothing too spectacular, but solid.
The real surprise came from a slightly different source: the public markets for quantum companies. We’re investors in IQM, the European leader in quantum computers, so we follow that space closely. Revenue multiples for listed quantum companies have exploded — in some cases, we’re talking 100 to even 2,000 times revenue. There’s one company with perhaps half a million to one million in revenues that has traded between €2 and €4 billion in market cap. It’s crazy — but a positive surprise if you’re a quantum investor.
Fundraising in 2025: Still a Hard Environment
Oleksandr: – Another thing you said in 2024 was that 2025 wouldn’t be easier for VCs raising new funds. Were you right?
Tom: – Unfortunately, yes. I don’t have the full global statistics, but from what I’ve seen and heard, fundraising remained tough throughout 2025.
A lot of mid-sized, non-mega-brand funds — serious, professional VCs but not the top-three names on everybody’s list — struggled. Many didn’t raise new funds at all, or raised smaller funds than they had hoped.
We did manage to close our fourth core fund about two months ago at €105 million, but that was below our target. In Q2, we had several LPs deep in our second-level data room. Typically, at least half of them would convert. This time, all five walked away. That left us with a smaller fund than planned, and it was a surprise.
So yes, I’d say I was right: 2025 was still a very challenging fundraising year for many VCs.
Going Earlier Into AI – and Five New Deals
Oleksandr: – You also said you were shifting OpenOcean’s strategy to invest a bit earlier: rounds of €3–5 million with tickets of around €2–2.5 million. How has that played out? Has it given you access to the hottest deals, as intended?
Tom: – I wouldn’t say we faced unexpected challenges, but the AI market has become extremely fast and competitive.
Friends of mine in the US — some of the top investors there — tell me about term sheets closing in a single day. There are literally just hours to issue a term sheet. When that’s the case, you must already know the founders, the space and your own thesis very well. Otherwise, you’re just gambling.
Post-COVID, deal cycles went from weeks to days, and now sometimes to hours. I haven’t seen that exact speed in Europe yet, but it’s definitely more aggressive and compressed than it was a few years ago.
For us, going earlier has worked as intended. We’ve kept good volume, and we’ve gotten into most of the cases we really wanted. Are they definitely the best deals? Time will tell, as always.
Oleksandr: – How many startups did you invest in during 2025?
Tom: – This year so far we invested in five companies:
- DreamFold – a super-AI company generating molecules (January)
- Coherent Healthcare (February)
- Arbio (April)
- Levels (May)
- Minerva (around August) – super fast growing AI for public tender automation
So we’ve increased the pace a bit, exactly as planned, with the slightly smaller tickets.
DreamFold is particularly interesting to me. If they manage to generate truly revolutionary proteins — which is the goal — they have a technology that scales better than Google DeepMind’s AlphaFold, and they can tackle more complex molecules. Whether they become as big as Isomorphic Labs, which raised around $600 million in its last round, remains to be seen. But I know they are already designing proteins others simply can’t. Technically, they’re world-class.
Profitability vs High-Growth Burn
Oleksandr: – In our last interview, you said that in 2024, profitability had become crucial, and that roughly 10 of the 15–20 companies per fund had reached it. Looking at 2025, how did that thesis play out? Are profitable startups winning, or are high-growth, cash-burning companies coming back?
Tom: – In AI — and also in quantum — profitability has become less central again.
If you look at OpenAI, Anthropic, or many quantum companies, revenues and burn are entirely out of balance. Some have huge revenues, some have modest ones, but almost all burn a lot of capital. These are very long-term bets, and the market is tolerating that because of the perceived potential.
So I’d say I was partly wrong. For most “normal” companies, especially outside those frontier areas, it’s still smart to be very careful with burn and not assume you can raise the next round easily. But if you’re in the hottest AI or quantum segments and growing fast, profitability isn’t the primary focus right now.
What Still Excites a 20-Year VC
Oleksandr: – After more than 20 years in venture, you’ve seen billion-dollar exits and companies fail. What actually keeps you up at night now — in an excited way, not worried?
Tom: – The same thing as always: working with exceptional entrepreneurs who are literally creating the future.
The future keeps coming, and as long as your mind stays somewhat sharp and somewhat young, it’s an incredible job. The themes change — from databases to SaaS to AI molecules and quantum — but the core is the same: being on the journey with extraordinary people.
Sometimes that’s about pure intelligence. With DreamFold’s founder, Max, for example, his brain just operates on a different level. He’s a Ukrainian physicist who went to MIT for his PhD, studied under a legendary Harvard Medical School professor who has launched dozens of successful startups from his lab, then moved to Mila in Montreal to work with Yoshua Bengio, the “father of deep learning”. Now he’s spun his bio-AI work out into DreamFold.
We have a powerful board: the former CEO of Bayer pharma, Alex from IQ Capital with a PhD in biochemistry, a serial molecule entrepreneur whose growing his second company already, and then me and another AI investor as observers. And even with that board, it’s hard to keep up with Max. That’s truly exciting.
Other times, it might be a brilliant communicator or a business genius, but in Max’s case, it’s deep technical, model-level genius. That’s the kind of thing that still excites me.
Beyond AI: Is Quantum the Next Wave?
Oleksandr: – Let’s go “beyond AI”. Based on what you’ve seen this year, are there technologies or business models emerging now that could define the 2027–2030 investment landscape?
Tom: – I think quantum could very well be the next big wave.
On the one hand, you have public-market behaviour: valuations for a handful of listed quantum companies have skyrocketed. On the other hand, you have real technical breakthroughs. Companies like Google and IBM are already demonstrating quantum advantage in particular, narrow tasks.
The major quantum players have also published clear roadmaps. When you follow them, it looks increasingly plausible that quantum computers will “really work” for meaningful use cases within a few years.
For us, this is obviously reinforced by our investment in IQM, the European quantum leader. They’ve raised €600 million in total, over €275 million in the last round, they’re now a unicorn, and they keep doing bigger and more complex things. Watching that from the inside gives you a lot of confidence that quantum isn’t just hype.
There’s also a fun circularity to it: quantum companies are using the best AI techniques to improve quantum algorithms and build hardware, and in the future, those quantum machines will be used to run AI algorithms much more efficiently. AI is helping quantum, and quantum will help AI.
Sam, Tony and the Next Generation at OpenOcean
Oleksandr: – In our 2024 interview, you mentioned that Sam and Tony had become partners at OpenOcean and were increasingly taking ownership of deals. How has that evolution continued over the last year?
Tom: – The main development is that they now truly own deals end-to-end.
Sam invested in Embeddable at the end of last year and then led the investments in Levels and Arbio. Tony led the investment in Coherent and, together with Patrik, in Minerva. They have led most of our new investments in 2025.
They are a crucial part of the firm’s next generation. They’re the ones who will keep OpenOcean evolving when people like me start spending more time on the golf course. They keep us moving fast and aligned with where the world is heading.
AI Roll-Ups and the Arbio Playbook
Oleksandr: – You briefly mentioned AI-enabled roll-ups. Could you explain the theme and how Arbio fits into it?
Tom: – About a year and a quarter ago, we started looking systematically at what we call AI-enabled services or AI roll-ups.
The idea is to take more traditional service businesses — often fairly “boring” ones — and radically improve them with an AI and automation stack you’ve built. Then you keep acquiring more such companies and roll them up under this AI-enabled platform and keep improving operations and margins.
We’ve made two investments in this theme so far:
- Arbio, which is a fantastic case
- Levels, which is at an earlier stage still
Arbio is a great example. They’ve acquired 27 or 28 service businesses in the holiday rental operator space. Historically, these companies had gross margins around 20–25%. Arbio applies its AI stack to increase revenues, cut costs and streamline operations. They need fewer people and run much more efficiently. On average, they’ve lifted gross margins to around 80% — essentially software-type margins.
On paper, that’s the dream case. And it also shows how fast the VC world moves: after doing two deals like this and a lot of research, suddenly you find yourself on some “dream team” market map as a midfielder, number 7 — which I think is the coolest position. It demonstrates how quickly you can become recognised in a new area when you move early and decisively.
2026 Outlook: AI Hype vs. Long-Term Wave
Oleksandr: – To wrap up: what’s your take for 2026? What do you think most people believe will happen, and where do you disagree?
Tom: – I don’t think there’s a single consensus, but many people now say AI is overhyped. I agree that there are bubbles and some strange, circular financing structures between big tech, cloud providers and AI companies.
But I don’t think the AI market is about to crash. Public markets always have some corrections, but I believe the general AI wave will continue for quite some time. We’re still early.
Our portfolio companies, startups that is, are already improving their operations by 20–40% using AI tools — just on internal efficiency. On top of that, they can sell better and more efficiently with an AI-supported go-to-market.
If fast-moving startups can do that, then big, slower enterprises have enormous room to improve. We’re still at the stage where, depending on which report you read, maybe 80% of companies have “tried” AI, but only about a third or less see real, measurable benefits at the EBIT level. That’s a sign that we’re still in the early innings.
So my view is that AI will keep driving improvements and value creation for years, and AI-focused companies — the ones that prove real ROI — will have a long runway ahead of them.
Oleksandr: – Then let’s see in November 2026 how your 2026 predictions play out versus reality. Thank you so much, Tom.
Tom: – Thank you. These interviews are always valuable. You end up articulating things you didn’t know you were thinking, which is helpful for me as well.
Oleksandr: – That’s precisely why I love doing them. They’re real dialogues, and it’s always interesting to see what picture emerges when all the pieces come together.
Tom: – Exactly. Thanks again, Oleksandr. Let’s catch up next year and see what the future has brought us this time.
0 comments