E27 -@DaveShap Confirms P(Doom) is 0%: Could All Machines Be Benevolent? | Humanity Unchained

In this episode of 'Humanity Unchained,' host Julia McCoy welcomes AI expert David Shapiro to discuss the alignment problem in artificial intelligence and the journey towards creating benevolent machines. Shapiro shares his intriguing experiments with GPT-2, his YouTube and AI career, and the development of his book 'Benevolent by Design.' The conversation covers the feasibility of achieving benevolent AGI, the current state of AI companies like Anthropic and OpenAI, and predictions on how AGI will reshape society. They also delve into the optimism surrounding AI's potential and the steps necessary to ensure that AI systems align with human values. Subscribe for insightful discussions on AI's future impact on humanity.

Follow Julia McCoy
‪@JuliaMcCoy‬

Follow David Shapiro
‪@DaveShap‬

00:00 The Unexpected Start of an AI Journey
01:15 Welcome to Humanity Unchained
02:02 Introducing David Shapiro
03:28 The Alignment Problem and Benevolent AI
06:53 The Three Heuristic Imperatives
09:53 Claude: The Benevolent Machine
14:44 The Future of AI Companies
19:46 Discussing the Book 'Who, Not How'
20:18 The Mystery of the Strawberry Account
21:37 OpenAI's Hype and Delivery
22:28 Advancements in AI Voice Technology
25:38 The Future of Humanoid Robots
27:22 Optimism in AI and Job Automation
28:11 Global Cooperation and AI Safety
35:01 The Absence of Malevolent AI Evidence
37:33 Closing Thoughts and Future Topics