Interview, Fireside Chat
a16z Podcast | Making Sense of Big Data, Machine Learning, and Deep Learning
- Big data is redefined by its utility for machine learning rather than the reverse, becoming qualitatively different once sufficient data exists for learning.
- Companies achieving a competitive edge will be those that are data intelligent and data driven, with Google historically treating its initial data batches as big data.
- Machine learning will transition from a standalone function to an inherent property of every application.
- Future products will incorporate deep learning capabilities, while internal machines are already communicating with each other at accelerating speeds.
- Anticipated outcomes include systems with "negative latency" that predict user needs, interactive learning capabilities that mimic human intelligence, and the boosting of human intelligence through machine intelligence in the near future.
- The drive to explore and build intelligence is described as an inevitable human pursuit, even as the resulting intelligence may differ from human intelligence.