March 4

Personal

  • Met up with family in new jersey, they have a cat and i love cats—i left quickly after stepping on its tail Work/Projects
  • TheoremExplainAgent is public!
  • Unstuck
    • Landing page done
    • Need to build MVP out, truly an execution issue right now
    • Talking to a lot of people—positioning it as a customer service tool first
      • We have a design partner. I’m thinking we will only do 1 of these along with close friends that want to try unstuck
  • Started doing some nerf/gaussian splatting stuff. Trying to live put my space in one iphone into another space through another iphone. Finally bringing out the vision pro
  • Hackathon
    • Things are moving very slowly, waiting for VCs to figure stuff out. it’s nice because they are taking some burden but it’s sooo slow Thoughts
  • A little pissed about the interview coder thing—had right vision at right time and just didn’t execute

Feb 18 2025

Personal

  • nothing Work/Projects
  • Unstuck hackathon - we won :)
    • thoughts on making it a product
  • wow—very little progress this week, not sure what happened. Lots of work work ig Thoughts
  • Lock in

Feb 12 - 17 2025

Personal

  • Got back into cubing Work/Projects
  • Robot arms
    • Securing funding for hackathon
    • Implemented IK for the arms and almost have vision based teleoperation working
    • Setting up mujoco env
  • Manim video gen paper submitted!

Feb 4 - Feb 11 2025

Personal

  • nothing Work/Projects
  • Video Gen eval paper
    • need to try some other methods for evaluation—deadline coming up
  • Laundry folding Robot
    • Robot moving/teleoperating
    • Also started video teleoperation module and data collection
    • ”launched” hackathon on twitter and getting good inbound from participants and
  • Starting a new position at robotics research lab at uwashington / allen institute! mentor is very very good—now just have to put the work in. Thoughts
  • happy with more focused work

Jan 28 - Feb 3 2025

Personal

  • Nothing really Work/Projects
  • I have to practice more agency at work. My job is less technical and more spelunking through people/docs. It’s def a muscle that I haven’t used much before but I don’t feel the impact just yet
  • Autonomous Hair Clippers
    • minimal progress here
  • Technical interview helper
    • no progress
  • Video Gen eval paper
    • Doing well, on to human annotation now
    • Had an interesting idea for using LLMs as a judge: finding the vectors of questions that are asked such that they are orthogonal to eachother, meaning they are completely separate from one-another
  • Laundry Folding Robot
    • Went to betaworks, and hackathon is coming along
    • Arms are built! Facing a bug for teleoperation (a motor might be broken)
  • Pitch Perfect
    • Small app to train pitch—once again got to prototype but no launch
  • Thoughts
    • I’m really stuck on what to work on long term. I’ve been doing a lot of thinking for this. I don’t really like the exploring stage, because it is doing a bunch of cool things but I feel like my strength is generally tenacity/grit/resourcefulness which ends up being useful when I do one thing for a long time
    • Because I have so many things going on, I notice I divert when things get hard, but I get stuck in an infinite loop.

Jan 21 - 27 2025

Personal

  • Hosted my first party Work/Projects
  • Learning to deal with bureaucracy. It’s difficult for me to work on things that I don’t think are important, but other people do.
  • Autonomous Hair Clippers
    • Got Merge Grant!
    • No other progress :(
  • Technical interview helper
    • No progress :(
  • Video Gen eval paper
    • Progressing but I don’t think it’s an insane paper. Once again, I’m finding it difficult to put in the time required to publish a paper if I don’t think it’s anything crazy
  • Laundry Folding Robot
    • Had more calls for robot arm hackathon, but pretty much only verbal offers
    • Printed leader and follower arms, assembly left
  • Ghostwriter
    • Loaded language model in browser for inference Thoughts
  • Context switching is hard—losing time like this
  • I find myself switching contexts when I get to a hard part. It’s almost like naturally i’m averting away from the cumbersome or unclear parts of things. Will work on fixing this

Jan 6 - 20 2025

Personal

  • Moved into homebrew/residency Work/Projects
  • Started working on ML Infra/Research at X/xAI
    • Working on making the recommendation alg better from an infra side
  • Autonomous Hair Clippers
    • Did interview for merge grant today
    • Made iterations on design/path forward
  • Technical interview helper
    • Started development on this with Kesu
    • Landing Page
  • Video gen eval paper
    • gearing up for publishing
  • Laundry folding robot
    • Started setting up robot arm hackathon (March 22)
      • Struggling to close sponsors
      • Raise Robotics - Gary
      • Toggle Robotics -
      • Monumental Labs - Micah
    • ordered parts for first arm—waiting for motors
    • Set up residency hardware lab

Blockers

  • Yapped too much—need to start doing stuff