newsfilter.io
Conference Presentation, Fireside Chat, Interview

Inception Fund & AI: From Inception to Exit Building AI Giants from Europe and the Nordics

Founding, Growth, and Exit of Silo AI

  • Silo AI was founded in 2017 by Peter Sarlin, a former professor, to operate as a private AI lab without initial venture capital funding.
  • The company reached an exit value exceeding $700 million when acquired by AMD, representing the largest pure-play AI exit in Europe and surpassing Google's acquisition of DeepMind.
  • Silo AI operated on a bootstrapped model for five years, growing by hiring directly from the academic sector, with the majority of its workforce consisting of PhDs in AI-related fields.
  • By 2022, the company employed approximately 220 AI scientists and engineers before accepting strategic capital and eventually selling to AMD to amplify its global impact.
  • Sarlin adopted a contrarian, "boring" philosophy regarding AI, defining it as function approximators and A-to-B mappings rather than pursuing hype-driven capabilities.
  • The company's product strategy evolved from autonomous vehicle technology (2017–2018) to optimizing mobile image quality, medical diagnostics, and large language models (LLMs) based on transformers.
  • Sarlin views AI development as a linear progression over five to six decades rather than a sudden paradigm shift, though he acknowledges that engineering efforts during hype waves have produced unpredictable performance jumps.

Strategic Outlook for European and Nordic Startups

  • Europe lags behind the US in scaling technology companies due to the absence of a unified single market, hindering the ability to build billion-dollar enterprises.
  • European startups possess strong talent, access to compute, and the ability to attract global capital, but must bypass fragmented national markets to scale effectively.
  • Current opportunities lie primarily in the application layer, where companies like Lovable are demonstrating global competitiveness from day one.
  • The European ecosystem lacks a 100-billion-dollar tech company created in the last five decades, with the exception of Spotify crossing the threshold post-Draghi report.
  • Founders are increasingly adopting a "global-first" mentality, targeting the US market immediately despite being headquartered in Europe.
  • Europe possesses distinct advantages in industrial platforms, including life sciences, automotive, robotics, and gaming, which are ripe for transformation by state-of-the-art AI models.
  • The prevailing concern is whether European industrial incumbents can leverage AI to maintain market position or if US tech giants will dominate these sectors.

Infrastructure, Sovereignty, and the "AI Rails" Debate

  • Europe missed the opportunity to create foundational mobile operating systems during the mobile shift (e.g., failed Symbian, limited Linux success) and faces similar challenges in the AI platform shift.
  • Potential European players in the foundational layer, such as Mistral AI, face significant hurdles in matching US distribution capabilities without equivalent platform dominance.
  • Sarlin argues that AI sovereignty requires more than just compute infrastructure; it necessitates a complete software stack and controlled distribution channels.
  • The "AI Gigafactory" initiative aims to establish European hyperscalers and AI clouds to reduce dependency on single suppliers.
  • Silicon diversity is identified as a critical component of sovereignty, with a strategic push to avoid over-reliance on NVIDIA GPUs and incorporate alternatives like AMD.
  • Successful sovereignty depends on allocating compute resources to the right business models and initiatives rather than merely building physical infrastructure.

The Primary Barrier to Scale

  • The single most critical barrier preventing the creation of European trillion-dollar companies is the lack of a functioning single European market.
  • Fragmented national legislations and languages force startups to build separate go-to-market strategies, creating inefficiencies that make the US market (one customer vs. 27 fragmented ones) the only logical path for scaling.
  • Initiatives such as "EU Inc" and the "28th regime" pushed by Ursula von der Leyen aim to address this fragmentation, though their tangible impact remains to be determined.
  • The consensus among current founders is that scaling within Europe is currently impossible without externalizing to the global market.