Jack Morris on Finding the Next Big AI Breakthrough

In this episode of the Odd Lots podcast, hosts Joe Weisenthal and Tracy Alloway interview Jack Morris, an AI researcher, to discuss the current state of AI research, model evaluation, and the factors driving progress in the field. They delve into the differences between public and private AI research, the challenges of measuring AI improvement, and the importance of data sets. The conversation covers topics such as supervised versus reinforcement learning, the potential of personalization and online learning, and the competitive landscape between US and Chinese AI labs. Morris shares his insights on the future of AI, the balance between scientific curiosity and monetization, and the ethical considerations surrounding AI development.

Outlines

Part 1: AI Landscape and Evaluation

Part 2: Training, Data, and Open Source

Part 3: Future Trends and Impact

Part 4: Summary and Outlook

Sign in to continue reading, translating and more.

Open full episode in Podwise

Odd Lots

Part 1: AI Landscape and Evaluation

Introduction to AI Research and its Current State

The Role of AI Researchers and Measuring AI Improvement

Challenges in AI Evaluation and the Quest for Human-Like Intelligence

Part 2: Training, Data, and Open Source

Supervised vs. Reinforcement Learning and the Slowdown in AI Progress

Open Source vs. Closed Source AI and the Importance of Data

Collaboration in the AI Ecosystem and the Advantage of Chinese AI Labs

Part 3: Future Trends and Impact

DeepSeek's Impact and the Sustainability of the US AI Lead

The Future of AI: Personalization, Online Learning, and Economic Impact

Open Source AI Downsides, Model Training Data, and Career Choices

Part 4: Summary and Outlook

Concluding Thoughts on AI Research and Future Directions

Jack Morris on Finding the Next Big AI Breakthrough

Odd Lots

Part 1: AI Landscape and Evaluation

00:48Introduction to AI Research and its Current State

Introduction to AI Research and its Current State

03:30The Role of AI Researchers and Measuring AI Improvement

The Role of AI Researchers and Measuring AI Improvement

07:28Challenges in AI Evaluation and the Quest for Human-Like Intelligence

Challenges in AI Evaluation and the Quest for Human-Like Intelligence

Part 2: Training, Data, and Open Source

12:36Supervised vs. Reinforcement Learning and the Slowdown in AI Progress

Supervised vs. Reinforcement Learning and the Slowdown in AI Progress

19:15Open Source vs. Closed Source AI and the Importance of Data

Open Source vs. Closed Source AI and the Importance of Data

24:02Collaboration in the AI Ecosystem and the Advantage of Chinese AI Labs

Collaboration in the AI Ecosystem and the Advantage of Chinese AI Labs

Part 3: Future Trends and Impact

27:46DeepSeek's Impact and the Sustainability of the US AI Lead

DeepSeek's Impact and the Sustainability of the US AI Lead

32:13The Future of AI: Personalization, Online Learning, and Economic Impact

The Future of AI: Personalization, Online Learning, and Economic Impact

37:37Open Source AI Downsides, Model Training Data, and Career Choices

Open Source AI Downsides, Model Training Data, and Career Choices

Part 4: Summary and Outlook

44:01Concluding Thoughts on AI Research and Future Directions

Concluding Thoughts on AI Research and Future Directions