At Rise of AI 2023 in Berlin, a panel with Tina Klüwer (director of KIEZ, the Berlin universities' AI entrepreneurship center), Alexandra Beckstein (co-founder and CEO of QAI Ventures), Thomas Neubert (founder of the Transatlantic AI eXchange, formerly at Intel) and moderator Johannes Otterbach (VP at Merantix Momentum) discusses how to build and grow AI ecosystems in Europe. Klüwer contrasts the individual, deep-dive culture of academic research with the fast, collaborative but shallower startup ecosystem, and argues both need exchange to succeed. Beckstein describes building a quantum and AI hub in Basel, Switzerland, where her venture fund gives startups corporate-partner projects, mentoring and hardware access, and flags that Europe has only two quantum-focused VC firms and just one fund investing in AI startups from Series C onward. Neubert traces his experience bringing Silicon Valley graphics technology to market and building the Transatlantic AI eXchange from a hobby into a network of over 8,500 followers and 25 events in two years. The panel discusses Europe's market fragmentation, the need to attract and retain talent, the balance between innovation and regulation, and the cultural mindset shift needed for European AI founders to take more risks.
Key takeaways
- Klüwer says academic ecosystems favor deep individual work while startup ecosystems favor fast, shallower collaboration, but both need exchange to succeed.
- Beckstein says Europe has only two VC firms investing in pure quantum technology and just one fund investing in AI startups from Series C onward.
- Neubert built the Transatlantic AI eXchange from a free hobby project into a network of over 8,500 followers and 25 events in two years.
- Neubert argues Europe will likely never be as unified as the US given its history, so the goal should be maximizing output within its fragmentation.
- Klüwer argues public discourse focuses too much on AI's risks and too little on its potential, discouraging young founders.
Detailed notes
Academic versus startup ecosystems
Tina Klüwer, who has founded and exited several AI companies and now leads KIEZ, the Berlin universities' AI entrepreneurship network, contrasts academic and startup ecosystems from her own experience straddling both. She says academic work tends to be individual, with small groups digging deep into a topic over years, while startup culture is fast-moving and collaborative but less deep, with people constantly exchanging ideas. She argues that exchange, interaction and feedback from peers are the crucial ingredient both ecosystems need to succeed, and that concrete shared tasks, rather than abstract goals, are the biggest motivator for groups working together.
Venture capital as an ecosystem magnet
Alexandra Beckstein describes how her firm QAI Ventures built a quantum-and-AI hub in Basel, Switzerland, giving startups hardware access to industry. She says that before QAI Ventures had a fund, startups would look around, connect with universities or corporate partners, and leave again; once the fund existed, it became a magnet that keeps startups engaged through global corporate-partner projects, one-to-one mentoring, and free access to quantum and AI hardware. She cites one German startup now going through a five-month investment and mentoring process that has already produced three customers through QAI Ventures' network. She also warns that Europe has only two VC firms investing in pure quantum technology, and only one fund investing in AI startups from Series C onward, which she calls a real problem for raising bigger tickets, alongside regulatory uncertainty and university-held IP rights on startup spinouts.
Building the Transatlantic AI eXchange from a hobby to a network
Thomas Neubert, who moved to Silicon Valley from Germany decades ago and later spent six years at Intel leading AI incubation programs, recalls helping bring 3D graphics workstation technology to the PC market despite no existing software-vendor ecosystem, and says the lesson he has since reapplied is to keep things simple and always answer the question "what's in it for" every party involved, from VC to entrepreneur to channel partner. He describes starting the Transatlantic AI eXchange two years ago as a free hobby project connecting Silicon Valley AI experts with European research institutions, universities, startups and incubation programs, inspired partly by a paper called "AI Made in Germany." It has grown into hybrid events in London, Munich, Berlin and Düsseldorf, with over 8,500 followers and 25 events reaching more than 120 experts, all run for free.
Fragmentation, talent retention and raising bigger funds
The panel discusses why Europe's fragmented national, language and regulatory borders slow ecosystem building compared with the United States' single government, language and huge capital base. Klüwer says Europe must stop losing researchers who leave to continue their work elsewhere and needs to attract and retain both young and experienced talent. Neubert argues Europe will likely never be as unified as the United States given a thousand years of separate history, so the practical question is how to maximize output within that fragmented environment; he also praises the energy he now sees in Berlin's incubators, comparing it to Silicon Valley 10 to 15 years ago, and cites a forecast of 27 percent annual growth in the generative AI market over the next decade as reason to attract more capital.
Sovereignty, regulation and the risk-versus-potential debate
On European technological sovereignty and AI regulation, Klüwer argues an ecosystem should represent all its stakeholders' perspectives, including startups, universities and large enterprises, and bring those views to decision makers, while warning against staying siloed in national "gardens" instead of collaborating globally. Neubert says he has shifted his own view after attending sessions at the European Commission's new San Francisco office, and now believes Europe's value-driven, privacy-focused approach to regulation is well positioned to set a global standard, noting that historically it was European regulators, not American ones, who fined Microsoft and Google for misbehavior. Klüwer argues public discourse focuses too heavily on AI's risks and too little on its potential, and calls for more visible role models and a stronger culture of tolerating failure to encourage young European founders.
Q&A highlights
Asked where Europe's entrepreneurial mindset shift is happening fastest, Neubert says he cannot yet point to a specific unicorn but describes the spirit and energy he now sees in incubators such as Merantix's AI campus in Berlin as comparable to Silicon Valley years ago, while cautioning that the capital infrastructure to take startups to the next funding round is still missing. Closing on what each panelist wants to see in the next 12 months, Klüwer expects more applications built on foundation and large language models; Neubert hopes for a shared Western AI regulatory framework and five to ten new European AI unicorns; and Beckstein wants to see deeper two-way convergence between AI and quantum computing.
Quotes
If you keep it simple, you have to simply answer the question: what's in it for?
Thomas Neubert
There is only one fund that invests from Series C on in Europe into AI startups, and I think this is a real problem.
Alexandra Beckstein
I'm getting asked a lot about the risks of AI, and I almost never get asked about the potential of AI.
Tina Klüwer
Questions answered in this talk
How many VC funds invest in AI startups from Series C onward in Europe?
According to Alexandra Beckstein at Rise of AI 2023, only one fund in Europe invests in AI startups from Series C onward, which she calls a real problem for raising larger funding rounds.
What makes a healthy startup and academic ecosystem, according to Tina Klüwer?
According to Tina Klüwer at Rise of AI 2023, both academic and startup ecosystems need constant exchange, interaction and feedback among peers, even though academic work tends to be individual and deep while startup work is fast and collaborative but shallower.
Why is Europe's AI ecosystem more fragmented than that of the United States?
According to Thomas Neubert at Rise of AI 2023, Europe lacks the single government, single language and unified capital base the US has, and instead has separate national, cultural and legal systems built over a thousand years of history, which he doubts will fully unify.