m.e.driscoll: data utopian: what to feed the mythical machine learning beast? →
One of the holy grails of machine learning is the creation of a system that can “read the web” and learn from it, much the same way that Isaac Newton picked up Euclid’s Elements in a Cambridge bookshop and taught himself geometry.
Imagine a mythical beast that could speed-read one-hundred…
But I wonder if some of our challenges in creating this mythical learning machine lie with what we’re trying to feed the beast. After all, the web of documents was written for human consumption. Natural language is a lossy compression algorithm; it maps the massive varieties of our experiences into semantic text. A high-frequency sensory stream of sights, sounds, and experiences gets hashed into “sidewalks can be slippery when it’s cold.”
To that end, if we want to create machines that can learn and reason about the world we live in, we should stop trying to give them our digested cud of content. We should provide them with direct experience, via the data streams that our instrumented planet is emitting from a growing network of sensors: weather stations, transit networks, electrical grids, smart phones, fitbits, and GPS devices. With that information, machines might begin to intuit relationships between weather and sidewalk slips — in forms that are beyond our own human minds to comprehend.
