In this episode, Marcel Salathé talks with Babak Falsafi, professor at EPFL and founder of EcoCloud, about the growing energy demands of AI and cloud computing. They dive into the real-world impact of data centers, the technological and geopolitical dynamics of AI infrastructure, and what it means for sustainability. From GPU power trends to cooling challenges, and from energy-efficient algorithms to national strategies, this conversation unpacks the complex trade-offs at the intersection of AI and energy.
Find out more :
https://parsa.epfl.ch/~falsafi/#portfolio-area
https://www.sdea.ch/
--------
56:52
--------
56:52
Conversation with Stef van Grieken, Cradle
In this episode of Inside AI, Marcel sits down with Stef van Grieken, co-founder and CEO of Cradle, the company rewriting the rules of biology through generative AI.
From a small Dutch village to building Google’s TPU and now the frontier of synthetic biology, Stef shares the journey that led him to ask a transformative question: What if we could design proteins the way we generate images or text? Cradle’s software is already helping design next-generation drugs, materials, and sustainable alternatives to oil- and animal-based products — even enabling interns with no biology background to outperform billion-dollar pharma solutions.
--------
1:14:36
--------
1:14:36
Conversation with Anton Osika, Lovable
In this episode, Marcel speaks with Anton Osika, CEO and co-founder of Lovable, the fastest-growing AI startup (by ARR) in 2025. Anton shares insights into Lovable's journey, exploring how recent advancements in large language models (LLMs) have enabled their rapid takeoff. He highlights current technical challenges, including growth strategies and navigating security issues inherent in AI-driven coding, and also provides a glimpse into Lovable's roadmap. Finally, Anton reflects on the experience of launching and scaling Lovable within Europe's startup ecosystem.
--------
24:50
--------
24:50
Conversation with Nicholas Christakis, Yale University
In this mind-bending episode of Inside AI, we dive into the fascinating world of human-AI hybrid systems. Forget your sci-fi dystopias — Marcel and Nicholas Christakis talk about bots that make us better humans. From robots that crack jokes to boost group morale, octopus's consciouness, and digital assistants that might be raising your kids (politely or not), we explore how social AIs can change the way we interact — with each other and with machines.
Christakis also explains why your AI co-pilot should maybe come with flaws (on purpose), how simple bots can catalyze group cooperation, and why your emotional bond with ChatGPT might be more revealing than you think.
--------
1:00:00
--------
1:00:00
Conversation with Charlotte Bunne, EPFL
In this episode, Marcel speaks with Charlotte Bunne, professor at EPFL specializing in AI and molecular medicine. Charlotte's groundbreaking research focuses on using AI to tackle challenges in personalized medicine, particularly oncology. They discuss her work on developing AI systems inspired by biological structures, the creation of AI Virtual Cells that integrate multiple biological scales, and the concept of Digital Twins to enhance diagnostics and treatment. From better understanding disease mechanisms to predicting the efficacy of new drugs, Charlotte explains how AI is transforming medicine at a molecular level.
France