HBR IdeaCast HBR Presents: Exponential View with Azeem Azhar
Apr 18, 2019
Joanna Bryson, a leading researcher in artificial intelligence and ethics from the University of Bath, joins Azeem Azhar for a fascinating discussion. They dive into the urgent need for ethical standards as AI technology evolves. Bryson highlights the risks of bias in AI systems and the importance of human oversight. They also tackle the challenges of explainability and accountability in AI, advocating for transparency in technology. The conversation touches on the need for interdisciplinary research to ensure responsible AI development.
AI Snips
Chapters
Transcript
Episode notes
Defining Intelligence and AI
- Define intelligence as behaving appropriately to context.
- This broad definition includes AI as artifacts built by humans expressing intelligence.
Implicit vs. Explicit Behavior
- Most human behavior is implicit and driven by subconscious factors, not explicit goals.
- AI can mimic explicit goal-setting, but implicit behavior is a significant part of intelligence.
Smart Devices and Data
- Azeem Azhar's thermostat sent 60GB of data from his home network, highlighting data usage by smart devices.
- Bryson notes AI is more than math; it's computation requiring time, space, and energy, referencing data centers' consumption.