The Good Fight David Bau on How Artificial Intelligence Works
69 snips
Sep 30, 2025 David Bau, an Assistant Professor and expert on deep generative networks, joins Yascha Mounk to dive into the complexities of AI. They discuss the critical need for understanding behind AI technologies and the implications of many not grasping their workings. Bau clarifies the distinction between generative models and classifiers, explaining how neural networks are constructed. He highlights the transformative role of transformers and shared insights on training methods, emphasizing the importance of moral alignment in AI development.
AI Snips
Chapters
Transcript
Episode notes
Engineering Needs Transparency
- Modern ML has encouraged engineers to accept opaque black-box models instead of understanding systems deeply.
- This shift risks training computer scientists to avoid inspecting how their systems actually work.
LLMs Are Generative Language Models
- Large language models are generative: they model open-ended human language rather than narrow decisions.
- Imitating language gives them far richer behavior than traditional classifiers.
Limitations Of Classifier Shortcuts
- Classifiers focus on the most salient difference and can take shortcuts that fail on atypical examples.
- This makes them accurate but prevents them developing broad world understanding.