flâneur

Michael Frew

0 followers · 286 views

on the atlas — 35

highlights — 6

  • These heuristics work for me (so far) because I have a strong bias towards novelty and exploration. Maybe one day, I’ll get tired of exploration, and these heuristics will change. When that happens, that’ll be growth.
    Measuring personal growth
  • The hallmark of integrity is honesty, not consistency.
    Changing Your Mind - by Deb Liu - Perspectives
  • We raise this point because it is common for ML practitioners to interpret a loss function as a useful signal during training—”if the loss goes down, all is well.” In policy gradients, this intuition is wrong, and you should only care about average return. The loss function means nothing.
    Part 3: Intro to Policy Optimization — Spinning Up documentation
  • if you could take a year off to work on something that probably wouldn't be important but would be really interesting, what would it be?
    The Bus Ticket Theory of Genius
  • It's a mistake to think they were "laying the groundwork" for the discoveries they made later. There's too much intention in that metaphor. Like bus ticket collectors, they were doing it because they liked it.
    The Bus Ticket Theory of Genius
  • he’s created this immense vertical filing cabinet in his brain of layers and layers and layers of files of information that he can draw back on now for more than 70 years worth of data
    Curius / Onboarding