🎧 This week on Unbreakable Ventures
This episode features Greg Headley, Lead Data Scientist at Informed Solutions, on what it actually takes to build AI systems that work in high-impact, safety-critical environments—from healthcare regulation to national policing.
Informed Solutions is one of the first ten UK organisations to achieve ISO 42001 certification, the world’s first international standard for responsible AI. With two Queen’s Awards for Innovation and clients including the UK National Health Service, Medicines and Healthcare products Regulatory Agency, Metropolitan Police, and Department for Environment, Food and Rural Affairs, they’re at the forefront of operationalising AI where public trust expectations are high.
In this episode, we cover:
- Why multi-disciplinary teams are essential for AI success
- The trust problem in AI—and how to solve it
- What the “95% of AI adds no value” statistic actually means
- Standards as innovation enablers, not barriers
- Digital twins for crisis preparedness
- The dark side of AI: cybersecurity concerns
- The next five years: scepticism, specialisation, and security
Key takeaway: Trust in AI is fragile and must be built incrementally. Organisations that invest in data governance, user-centered design, and responsible AI practices will be the ones who actually realise value from these technologies.
Greg’s recommendations:
Andrej Karpathy’s YouTube channel – AI education from a former Google/Tesla researcher
Dwarkesh Podcast – In-depth conversations including the Richard Sutton interview on reinforcement learning
Listen on Spotify, Substack, Apple, or YouTube.
This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.unbreakableventures.com