Research and Applied AI Summit: celebrating 10 years
Bringing you the best of AI today and what’s coming tomorrow.
For ten years now (!), our Research and Applied AI Summit (RAAIS) convenes a unique cross-section of the global AI ecosystem for a one-day deep dive into the science and real-world impact of artificial intelligence. The forum brings together researchers, engineers, founders, academics, and policymakers—each working to push the boundaries of what AI can do. We hope that the conversations that start here turn into collaborations, companies, research papers, and long-term friendships.
This summit is hosted by the RAAIS Foundation, a registered non-profit with a mission to advance education and research in AI for the common good. Every ticket and event helps fund open-source projects, new research initiatives, and educational tools that broaden access to cutting-edge AI.
We continue to be driven by two core beliefs. First, AI's impact will extend far beyond for-profit use cases. Open source and independent research initiatives are where orthogonal ideas are often born. Second, creating accessible pathways for talented contributors to get into AI—regardless of geography or background—is essential to shaping the future we all share.
The State of AI: 2025
This past year has seen an extraordinary acceleration in both AI research and real-world deployment. Frontier labs have released increasingly intelligent models—GPT-4.5, Claude 3, Gemini 2.5. We're now entering an era defined by reasoning and inference-time compute, with models like o3/o4, DeepSeek R1, and others still in the oven. These systems are powering a new generation of autonomous agents that can reason, plan, and act. The era of hallucinating chatbots is fading.
AI is now embedded in IDEs, operating systems, creative workflows, and enterprise dashboards. Personal copilots aren't just concepts—they're shipping. From text to voice to video, multimodal interfaces are rapidly becoming the norm. Embodied AI is also gaining traction, moving intelligence from the screen into the physical world.
Yet the open vs closed-source divide has widened. Open models are competitive at smaller scales but struggle for mainstream adoption. The most critical advances in scaling, tool use, and reasoning remain proprietary. Frontier labs and nation-states are responding with AI strategies, export controls, and public investments in compute infrastructure. Existential risk research has taken a back seat. The stakes have never been higher.
What to expect at RAAIS 2025
This year’s program explores the cutting edge of AI research and entrepreneurship—from frontier models and scientific discovery to creative tools, defense, and geopolitics.
We begin with Mati Staniszewski of ElevenLabs, whose work in real-time dubbing and lifelike voice interfaces is reshaping how we interact with machines.
Next, Andreas Blattmann of Black Forest Labs shares progress in generative video and 3D diffusion models that collapse creative workflows and unlock new modes of storytelling.
Our policy panel—Lionel Laurent (Bloomberg), Chris Yiu (Meta), and Benedict Macon-Cooney (Tony Blair Institute)—examines the tension between innovation and governance, from platform responsibility to global geopolitics.
Max Jaderberg of Isomorphic Labs brings us back to science, showing how AI is transforming the discovery and development of new medicines.
Eiso Kant of poolside explores frontier models purpose-built for software engineering, rethinking the developer workflow from the ground up.
Paige Bailey of Google DeepMind dives into model usability at scale—how to make advanced AI both accessible and safe in production.
Our defense panel—Dimitrios Kottas (Delian Alliance Industries) and Adam Satariano (The New York Times)—discusses how AI and autonomy are reshaping global power dynamics and what the West must do to protect liberal democracies.
We close with Edward Hughes of Google DeepMind, whose work on cooperative, adaptive agents tackles one of AI's grand challenges: learning and collaborating in the messy, open-ended real world.
What’s to come
At Air Street Capital, we continue to back AI-first founders at the earliest stages. The last 12 months have redefined what fast iteration and growth looks like. Agents, multimodal interfaces, and systems capable of reasoning are now flowing directly from research into production.
But the velocity of change brings volatility. Open-source models face mounting pressure. Frontier models are consolidating market power. Compute, data, and talent are now strategic assets—and geopolitical levers.
That’s why this community matters. RAAIS is where ideas get tested, where researchers meet builders, and where new ventures take root. The university spinout reforms we championed last year are being adopted across Europe, thanks in large part to many of you.
This year marks ten years of RAAIS. We’re proud of what this community has become: a space where rigorous research meets real-world impact. Thank you for being part of it.
Keep an eye out for the upcoming State of AI Report, launching this October, and thank you for reading Air Street Press for ongoing commentary at the intersection of research, industry, and geopolitics.
If you’re interested in providing financial, mentoring or operational support to The RAAIS Foundation, reach out to us at info@raais.org