Hi! i’m scott penberthy, or dr. scott to my friends.

  • i love science.
  • i like building cool things.
  • i enjoy teaching ai and space.

this site

i started scott.ai after ingesting NASA Science Mission Directorate’s (smd), “Astrophysics 30 year plan: Enduring Quests and Daring Visions.” smd’s mission is to “discover how the universe works, explore how it began and evolved, and search for life on planets around other stars.”

seriously. we have a government agency whose mission is star trek!

smd sketched an architecture to explore solar bodies, starting with the earth, then the moon, followed by mars, europa, and asteroids. i’m excited to be part of it!

the apollo mission relied on telemetry and humans. artificial intelligence and its autonomous robots, vehicles, prediction powers, and imagination powers will propel future missions.

lucky for me, i studied ai in school. i still do. this site is dedicated to things i’ll learn along our 30-year journey. i’ll share them with you.

my history

i was born four years before man walked on the moon. i spent my childhood obsessed with nasa, writing computer programs that simulated star trek games, apollo 11 launches, the lunar landar, and interstellar communications (online chat, before the internet).

i went to mit at 16 to become an astronaut. i failed out of mit rotc’s naval flight program after 2 months… my eyesight was imperfect. sigh. depressed and realizing my life was completely over (sounds like a teenager, doesn’t it?), i sulked across campus. there i ran into an open house for the mit artificial intelligence laboratory. i gave it a look.

i fell in love. i spent the next 4 years working in the ai lab, teaching, and building robots in the summer. many years later, my PhD would payoff. i started at google in 2016 doing ai fulltime, after a career of webly things.

a year later i met nasa.

i actually get to help nasa and their 30-year mission. i’m starting from within google, where i do “applied ai.” we model phenomena with tensor calculus and computational graphs, optimized through deep learning, reinforcement learning, generative adversarial networks, and more.