Executive Briefing on Artificial Intelligence and AI Agents with Prof. Marcel Salathé (Sparkr Podcast #35)

A lot of what is written about artificial intelligence today is more distraction than signal.

The strategically relevant questions are not primarily about the latest model release, voluntary agreements between tech executives, astronomical IPO expectations or predictions about the end of humanity.

What matters more is understanding the forces beneath the surface: which developments are likely to persist, how they may affect organisations over the coming months and years, and what they mean for strategy, capabilities and dependencies. In other words: What are the underlying principles and concepts that shape the AI era mid- to long-term?

An Executive Briefing on Artificial Intelligence

With exactly that goal in mind, I sat down with Prof. Marcel Salathé(opens in new tab) for the latest episode of the Sparkr Podcast.

Marcel is Co-Founder and Co-Director of the EPFL AI Center(opens in new tab), one of Switzerland’s most prominent voices on artificial intelligence, author of a bestselling book on AI(opens in new tab) and creator of “The Framework(opens in new tab)” – an app designed to help people understand AI more clearly and systematically.

Our conversation is intended as an Executive Briefing for leaders, strategists and decision-makers who want to engage with artificial intelligence compactly, but with sufficient depth.

We discuss questions such as:

• What remains strategically relevant when today’s leading AI models may already be outdated twelve months from now?
• Which underlying forces and principles are likely to shape AI over the next five years?
• Which assets, capabilities and dependencies determine whether an organisation becomes more productive through AI – or more vulnerable?
• What changes when AI moves from answering questions to autonomously performing tasks?
• Which concepts around LLMs, inference, agents and AGI should non-technical leaders actually understand?
• And how can executives distinguish meaningful technological change from hype, noise and alarmism?

A recurring theme throughout the conversation is that access to AI alone will not determine who benefits most. The strategic question is increasingly whether organisations are able to redesign workflows, develop judgement and capabilities around AI, and manage new dependencies on models, cloud infrastructure, chips and other parts of the AI value chain.