Innovation

• Some kind of innovation.
• Two types:
  • Gradual vs Radical
    • Gradual refers to light changes, or in other words, incremental improvements.
    • Radical refers to changing everything.
    • Good analogy:
      • Think of it as climbing a mountain.
      • Gradual is like taking one step at a time, always moving higher than where you currently are.
        • With this approach, you can reach a certain maximum altitude, but it may not be the best.
      • Radical is like parachuting down onto the mountain from above.
        • There are more chances of failure, but if successful, it can be innovative.
    • It feels like the gradient descent method in Machine Learning.
    • Can we apply techniques from information science to prevent getting stuck in a local optimum? (This kind of thing is interesting.)