"I want to give people the map ... what you do with it is up to you." Demba Ba has a 5-year ambition to cut open the AI Black Box and explain what makes LLMs work.

Demba Ba, the Harvard Professor described as an "AI Neurosurgeon", shows why explainability lies at the heart of humans taking control of AI's alien intelligence. And if we get this right then we will also gain huge insight - "a neurosurgeon's dream" - in to not only how AI but also humans think.

The conversartion ranges from the immediate opportunities from healthcare roll-out and AI audit (why did the AI make that decision?) to the potential power this could give the state to manipulate people. And if we can understand and predict AI behaviour, and do the same with people, what does this tell us about free will, or are we all just products of our training data? If we can create individual human knowledge graphs - building the human model - then does this mean we can predict an individual's actions? What are the implications for AI tweaking - does lobotomy (or AI dissection and editing) have a wider impact?

We also discussed the huge AI opportunity for Africa - potentially one of the biggest winners in the coming Age of Intelligence.

See omnystudio.com/listener for privacy information.

What can one learn from a Private Equity AI executive who quotes Leon Trotsky?

Francois led the Boston Consulting Group’s Henderson Institute in its recent impactful work on Generative AI and the workforce. Ideas such as the “jagged frontier” and “AI as an exo-skeleton” have helped create mental models for everyone working in the space.

He brings a unique perspective on AI adoption and competitiveness of companies and nations, combining decades of experience with BCG’s large clients and his new role working with SMEs in Private Equity. He argues that smaller companies have the advantage of speed and focus. He lays out a framework for how companies can embrace the technology - with a clear focus on how to bring people with them. This means firms must be ambitious with AI: they must “hunt elephants, not rabbits” and stay focused on the business case rather than any technical details that may not matter.

His experience working in China for years shows him the difference between countries with a sense of urgency and those who are too focused on defending existing structures. The key is to “protect individuals not job descriptions”, as he urges policy makers to do.

He provides insights on what children should study but reassures that not all jobs will disappear: “There will be consultants as long as there are learning curves”. Learning is at the core of intelligence and the only way to live in a world where … revolution is permanent.

See omnystudio.com/listener for privacy information.

Pattie Maes brings a unique perspective to the quest to build better machines: she pioneered the concept of the software agent in the 1990s. Ever since - as an entrepreneur and leading MIT academic - she has focused on the mission to help make humans smarter. To augment humanity, not robots.

She takes us on her professional journey - from her first company (Agents Inc) to her latest research at MIT. Pattie is focused on how computer systems and digital devices might augment people and assist them with issues such as memory, learning, decision making, health and wellbeing. In 2025 she was recognized with the "Lifetime Achievement Award in Human Computer Interaction" by the Association for Computing Machinery (ACM).

She talked us through her concerns about AI - the sheer power of a technoklogy that is increasingly intermediating between us and the world around us. But it is riddled with bias and prone to hijacking for often nefarious ends. And, as her research shows, humans are very open to simply falling in with this powerful persuasive force.

Pattie's words will be important for anyone wondering how to engage with this technology - wise warnings combined with a clear sense of all that should be possible if only we aimed for better. This raises the stakes for Europe - she argues that for reasons both selfish (national security) and idealistic (building a better world) Europe must aspire to take control of its "new utility".

See omnystudio.com/listener for privacy information.