Into the Impossible With Brian Keating Is AI Our PARTNER or Our ENEMY? Google CTO Blaise Agüera y Arcas
56 snips
Sep 30, 2025 
Blaise Agüera y Arcas, a senior researcher at Google and author, and Benjamin Bratton, a philosopher and theorist, dive into the evolving relationship between AI and humanity. They discuss whether AI can truly learn from us or if we’re the ones being trained. The duo explores AI's potential in enhancing education, medicine, and creativity, alongside the ethical implications of its use. They also debate the significance of randomness in human and machine creativity and propose the idea of 'universal basic compute' as a pathway to collective intelligence.
AI Snips
Chapters
Books
Transcript
Episode notes
Embodiment As Computation
- Blaise argues brains are fundamentally computational and encode sensations as spikes, so embodiment can be instantiated in other substrates.
- Virtual experiences can produce genuine embodied feelings, blurring strict human-only embodiment distinctions.
Hardware Lottery Shapes AI
- Blaise describes 'lock-in' where hardware and algorithms co-evolve, favoring solutions suited to existing tech like GPUs and transformers.
- He predicts hardware-driven trajectories and future shifts toward more energy-efficient architectures.
Use LLMs As Math Collaborators
- Use LLMs as collaborative tools for math and physics work by iterating with them and checking equations in LaTeX.
- Treat models as conversational partners to explore rearrangements, not as infallible final authorities.
