Peter Sarlin, who co-founded Silo AI, Europe’s largest private AI lab, is currently a Corporate Vice President at American semiconductor company AMD and head and co-founder of AMD Silo AI. He is also a Professor of Practice at Finland’s Aalto University, specializing in machine learning and AI. With a PhD in machine learning, Sarlin has a rich academic background that spans roles as a tenured professor, visiting professor and research associate at institutions such as Imperial College London, London School of Economics, University of Technology Sydney and Goethe University Frankfurt. Sarlin has also held board-level roles in IEEE Technical Committees and positions in early-stage startups as well as large organizations, such as the European Central Bank and the International Monetary Fund.
In addition to his work in academia and industry, Peter is the founder of Foundation PS, an initiative dedicated to advancing AI research and strengthening Europe’s AI ecosystem. The foundation has donated 13 professorships that form the PS Fellow network, and supports AI research through grants, scholarships, and other initiatives, ensuring collaboration between academia and industry to drive AI innovation forward.
Sarlin recently was awarded the Transformation Excellence Award 2025 by the digital transformation consultancy bluegain for his contribution to Silo AI and the wider AI community in Europe. The Innovator’s Editor-in-Chief conducted a fireside chat with Sarlin at the awards ceremony, which took place at bluegain’s CXO luncheon in Davos during the World Economic Forum’s annual meeting. He agreed to separately be interviewed by The Innovator about Silo AI, Europe’s position in the global AI market, and the upcoming global AI Action Summit in Paris.
Q: Why did you launch Silo AI?
PS: Most of my career was spent abroad but in 2017 I moved back to Finland. Instead of returning to a tenured professorship, I decided to expand the work that had been initiatied under Infolytika Ventures, a company that I founded in 2011. In 2017, I and my co-founders, tech veterans Tero Ojanperä, Johan Kronberg, Juha Hulkko, Kaj-Mikael Björk and Ville Hulkko, spun out Silo AI and aligned on a joint mission to build a private AI lab. Our mission was to work with Europe’s leading companies to help them remain competitive and ensure that the region would attract and retain AI talent, to build Europe’s flagship AI company.
Q: How has Silo AI’s focus evolved over time?
PS: Initially we served as an R&D partner of some of the biggest companies in Europe, working on some of the most advanced AI initiatives with Rolls Royce, Philips, Allianz, T-Mobile, Honda, and others. We didn’t raise venture capital. We bootstrapped until 2022. After we had together with customers grown to more than 200 AI researchers, we raised strategic capital to transition from being a private AI lab conducting strategic R&D partnerships to becoming a product business and releasing open source multilingual models trained on European languages. Today we build AI to enable smart devices, autonomous vehicles, industry 4.0, and smart cities and train state-of-the-art open source AI models. With advanced compute, a full-stack AI platform and world-leading AI scientists, our approach empowers organizations to develop AI that they own and control. Now obviously with a focus to ensure they succeed on AMD compute platforms.
Q: In July 2024 Silo AI’s $665 million acquisition by US-based AMD was the largest AI company acquisition in Europe‘s history. I have to admit my first reaction was’ Oh no another European company has been sold to the Americans. Why did you sell to AMD and why do you see it as a positive for Europe?
PS: It was a very deliberate decision. Silo AI did not have to sell and we were not actively looking to sell the company. My genuine thinking was ‘we can create even more value as part of AMD.’ Silo AI was never intended to be a platform play with hyper growth. We have always been a private AI lab. Our aim was long term sustainable growth. We were open to M&A and acquired a few companies, which had indeed fueled our growth during the past years. Eventually we decided thatjoining forces with AMD will make us stronger. We are still headquartered in Finland and can help European companies even more, while now obviously serving a global client base. This is a tremendous opportunity for Europe. AMD is committed to expanding our activities, enabling significant value to be created both in and outside Europe.
Q: On January 15 you announced a substantial donation to strengthen AI research in Finland, funding 13 professorships across Finnish universities and the ELLIS Institute under the auspices of your Foundation. You’ve publicly said that by funding professorships and supporting world-class research, the foundation aims to drive innovation, attract top talent, and solidify Europe’s position as a global leader in AI. Please tell us a little about your thinking on that.
PS: It is about giving back and wanting to ensure that Europe has world-class AI research. As a foundation, if we provide grants from profits, it is a slow process. What Europe needs, is a larger initiative with a long-term commitment, which is why I decided to launch a program that puts forward a significant number of professorships with permanent tenured positions and supports them in the long term. By concentrating on attracting the most talented professors, it’s expected to have a much larger impact, as they eventually attract research funding and recruit vibrant research groups.
In addition to the new professorships the ELLIS Institute [European Laboratory for Learning and Intelligent Systems Institute], a pan-European network established to advance fundamental AI research is officially initiating its activities in Finland with €40 million in funding from the Finnish government. Finland was also recently awarded Europe’s largest AI factory initiative, an ambitious undertaking supported by the European Commission with a total budget of €600 million. Europe already hosts LUMI, one of Europe’s most powerful supercomputers and a critical resource for AI research, with a unique compute quota reserved for Finnish research initiatives. This is the starting point to ensure that talent has exciting initiatives to join.
Academic positions, particularly PhD studentships and postdoctoral fellowships, are short-term jobs. This necessitates a constant global search for the next career step, highlighting the importance of world-class research and industry initiatives, such as these professorships, to retain top talent. Right now, the most talented researchers are joining leading tech companies to work on the most challenging problems with the most talented people. As in the Paris ecosystem, many of these companies might be US-owned, but by retaining talent, they eventually in a year or two contribute back, for instance by forming companies like H Company or Mistral AI. The next step Europe needs to focus on is strengthening hubs and the ecosystems in these hubs. While Paris is a vibrant AI ecosystem, Helsinki is increasingly becoming a strong AI hub.
Q: Some think it is already game over in the AI market and the U.S. and China have won. What do you see as Europe’s AI strengths particularly in light of DeepSeek’s recent breakthrough?
PS: Overall Europe needs to succeed in building scalable digital products and the tech companies that build and scale them. That is what Europe is missing. In 2025, I believe the world will see more and more AI agents and more specialized models, fine-tuned for specific industries, use-cases, or other domains, and products built on top of these models. We will also, and more importantly, see foundation models tackling new modalities such as life sciences, material sciences and robotics. We will continue to see significant innovation in all parts of the value chain, from compute to model architectures to applications. In that sense, the recent news of DeepSeek’s new open source models was not a big surprise. There is still a lot of opportunity for innovation, development, and growth in the AI space.
Q: Next week the French government is co-hosting the AI Action Summit, a global conference that will gather government and the sector’s principal actors. What do you see as the most important issues that need to be resolved to ensure that AI is developed responsibly and equitably, preserving privacy and protecting society and IP and what role do you think Europe can play?
PS: Europe has built strong foundations for AI development – from world-class research institutions to significant compute infrastructure, and increasingly interesting startups and scale-ups in the space. But its competitiveness hinges not on regulatory frameworks or standalone AI development, but on its ability to translate this potential into market-leading businesses at scale. We’re seeing promising examples emerge, with vibrant AI ecosystems developing in several European cities, but Europe needs to accelerate this momentum.
To lead in responsible AI development, Europe needs to focus on clear implementation of existing laws and relevant guidelines for all actors in the value chain, hopefully simplifying the regulatory landscape for scale-ups, encouraging public and private funding towards ambitious moonshot projects, and encouraging stronger links between academia and industry.
Q: In your role at AMD Silo AI you have worked with many corporates on their AI strategies. If you had to boil it down to just a few things, what advice would you give the companies to help them successfully integrate the technology into their companies?
PS: The McKinsey Global Institute talks about the tremendous real economic value potential of AI due to productivity and efficiency gains. My way of looking at this is that the winners are the companies that build AI into features of their products so they can grow and go after new revenue opportunities. The most successful companies will be those that have completed their digital transformation, embed both software and AI at the core of their business, and get better and better at what they can offer their customers through a data and AI flywheel.
To access more of The Innovator’s Interview Of The Week articles click here.
