Notes from a discussion of machine learning as a driving force with Andrew Lee. We discussed machine learning as a driving force of:

  1. The shift from products to services
  2. Increasing distrust of the media

Potential of Machine Learning / AI

  • Origins as a driving force:
    • Ability to solve problems that are difficult to specify or define
  • Historical precursors:
    • Gaming industry created cheap GPU’s
    • Mature tech companies with web-scale data sets
    • Foundational concepts such as back propagation
  • How does it influence the conditions which are influenced by it?
    • The shift from products to services
      • Companies have an incentive to create service platforms to collect data and collect as much data as possible
    • Distrust of media
      • Machine learning can synthesize media that is difficult for humans to detect as machine generated
      • Filters run by machine learning algorithms sometimes promote low quality information. eg facebook filtered feed, google suggested answers
      • Personalization can feel obtuse or “creepy”
  • What are the counter-forces at play?
    • Privacy regulation
    • Distrust of corporations
    • Fear of artificial intelligence and black box algorithms
    • Hype cycles / AI winters
    • Platform ecosystem stakeholders with competing interests. eg iOS not allowing certain tracking
  • Research on probability and random algorithms is the most popular field
  • Because such things as chess are no longer considered to be in the field of AI, but rather a better and faster performance of calling data.
  • In other words, they don’t require complex thinking process.
  • Although the definition of “intelligence” is still debatable, AI is approached in different direction.
  • Solving requests that require the understanding of contexts
  • Mathematics and statistics creating categories
  • Brain Simulation: the concept and scientific project of creating a computer-run model of brain neuron connections
  • Bottom-up approach: the piecing together of systems to give rise to more complex systems, thus making the original systems sub-systems of the emergent system (A top-down approach is essentially the breaking down of a system to gain insight into its compositional sub-systems in a reverse engineering fashion)
  • How to Create a Mind by Ray Kurzweil