Succeeding with AI
John Savill shares practical lessons from customer work and Microsoft’s own AI efforts, focusing on what drives real value, why initiatives stall, and how to think about the human side, trust, and risk when adopting AI.
Overview
The video is a practitioner-focused talk on succeeding with AI initiatives, covering adoption patterns, where AI can drive benefit, what makes an initiative “good”, and common places organizations get stuck.
Key themes include:
- AI adoption: how organizations approach adopting AI and what tends to accelerate or slow progress.
- Where AI can drive benefit: identifying areas where AI can create meaningful outcomes.
- What makes a good AI initiative: characteristics of initiatives that are more likely to succeed.
- Human and AI: how people and AI systems work together, and how to “amplify the human” rather than replace.
- Where organizations get stuck: typical blockers that prevent progress.
- Rethinking the organization: changes needed in structure and ways of working to support AI.
- Intelligence and trust:
- Intelligence: considerations around capability and how intelligence is applied.
- Trust: considerations around trustworthiness and confidence in outcomes.
- Models: discussion of models as a component of AI solutions.
- Bad actors using AI: risks and misuse scenarios that need to be considered.