{"type":"rich","version":"1.0","provider_name":"Transistor","provider_url":"https://transistor.fm","author_name":"Healthy Spaces","title":"AI Lab: The Feedback Loop Behind Trustworthy AI","html":"<iframe width=\"100%\" height=\"180\" frameborder=\"no\" scrolling=\"no\" seamless src=\"https://share.transistor.fm/e/7d93c59d\"></iframe>","width":"100%","height":180,"duration":1685,"description":"How do you build AI that can be trusted in the real world? \n  \nFor AI to move from research into practical applications, the lab and the field need to work together. Real-world deployment creates new insights for researchers, while research and testing help strengthen the technologies being put to work. In this episode, we explore how that continuous feedback loop can make trust a foundation for responsible AI development and deployment. \n  \nWe hear from Foutse Khomh, Vice President of Research and Innovation, Polytechnique Montréal, about what it takes to build trustworthy AI. He explores how researchers and industry can work together to understand how AI systems behave, identify their limitations, test them in realistic conditions, and develop the safeguards and controls needed for reliable deployment. He explains why ongoing collaboration between the lab and the field is essential to continually improving AI systems. \n  \nRiaz Raihan, Chief Digital Officer at Trane Technologies, takes us inside the BrainBox AI Trane Technologies AI Lab in Montreal, where researchers, engineers, product managers and building experts are developing practical AI applications for the HVAC industry. He discusses how the AI Lab connects research with real-world deployment, from autonomous building controls to new approaches to optimizing chilled-water systems, and how Trane approaches trust through security by design, compliance and partnerships. Together, they show how the ongoing exchange between research and real-world application can help build AI that is trustworthy, responsible, and capable of delivering measurable value.\n \nEPISODE CHAPTERS   \n00:00 Introduction - Building trust in AI \n01:00 What makes AI trustworthy? \n02:15 Understanding AI’s weaknesses \n03:45 Testing AI for the real world \n05:48 Building trust without overtrusting AI \n07:45 The Swiss cheese approach to trustworthy AI \n09:06 Why the lab and field need each other \n12:15 Inside the BrainBox AI Trane Technologies...","thumbnail_url":"https://img.transistorcdn.com/EWG6ZnM49FIe0Yy22Amb1KSChry1WZmpnrrC9-UXzGI/rs:fill:0:0:1/w:400/h:400/q:60/mb:500000/aHR0cHM6Ly9pbWct/dXBsb2FkLXByb2R1/Y3Rpb24udHJhbnNp/c3Rvci5mbS9zaG93/LzE2Mzc1LzE2ODYw/MzY2MDctYXJ0d29y/ay5qcGc.webp","thumbnail_width":300,"thumbnail_height":300}