• different stages of ai
    • imagenet era - + alexnet, fei fei li bets on collecting a huge dataset of images and when when we realized data labelling was the unlock towards..
    • deep learning era - generation before me,
    • generative ai / attention mechanisms era - being able to apply the compute and use transformers to accelerate understanding
  • diff areas of interest
    • computer vision
    • spatial intelligence (fei fei li world labs),

anecdotes from work / activities

  • i’ve always be interested in physical world & data visualizaitons

    • first research project in earth sciences, nerded out in high school, love heat maps
  • i’ve always been interested in nlp and linguistics

    • interesting problem in imagenet / cv dataset collection era where it’s up to the human to determine the ontology - how to label datasets and what categories neatly define an object’s characteristics
  • why is no one in ai4all off to a great startup, but so many ppl from mtsi are?

    • its bc mtsi has the audacity and will just do things
  • gnns