Machine Learning Street Talk (MLST) #64 Prof. Gary Marcus 3.0
Feb 24, 2022
In this engaging conversation, cognitive scientist Gary Marcus, founder of Robust AI, tackles profound questions about AI and consciousness. He expresses skepticism about AI systems claiming self-awareness and the philosophical debates around consciousness. The discussion dives into the challenges of abstract models and the importance of stable symbolic representations, particularly in self-driving technology. Marcus also reveals insights on extrapolation in high dimensions and scaling laws, emphasizing the complexities of true understanding in neural networks.
AI Snips
Chapters
Books
Transcript
Episode notes
LLMs and Consciousness
- Current large language models (LLMs) lack semantic grounding, reasoning, and coherence.
- They primarily perform statistical operations on data without true understanding.
GPS vs. LLM Self-Awareness
- Gary Marcus argues a GPS system exhibits more self-awareness than LLMs.
- It uses its location awareness for real-world decision-making, unlike GPT-3.
Qualia and LLMs
- Qualia, subjective experiences like feeling red, remain largely mysterious.
- LLMs, lacking internal states and consistent world models, likely don't experience qualia.
