Hello World šŸ‘‹

Iā€™m Andrei Rotenstein

  • šŸ”­ Iā€™m currently working on getting better at GPU programming and C++20.
  • šŸŒ± Iā€™m currently learning neurosymbolic AI and practical deep learning.
  • šŸ‘Æ Iā€™m looking to collaborate on some FOSS project where I get to do the above.
  • šŸ¤” Iā€™m looking for help with the previous bullet.
  • šŸ’¬ Ask me about CHRY 105.5 FM
  • šŸ“« How to reach me: rotenstein.andrei.v@gmail.com
  • šŸ˜„ Pronouns: he/him.
  • āš” Fun fact: When I was 12, Iā€™ve destroyed my Commodore 64 when I tried to use the Joystick port to drive motors.

Current Goals

Iā€™ve finally taken the time to set up this GitHub account, after much procrastination, with the intention of creating repositories to support personal exploration into several topics, as well as a GitHub Pages site documenting it all. The topics are:

  • Neurosymbolic AI. Iā€™m fascinated by what so-called foundational models cannot do. They cannot plan or reason (formally, in the ā€œtraditional AIā€ sense), contrary to popular opinion. There are lots of efforts in the research community to imbue models with the missing capabilities. And Iā€™m here for it.
  • C++ 20. I taught myself C++ when I was a kid, and so I have always loved that language. I also hate it. (Our relationshipā€™s complicated.) Those damn ISOC++ working group(s) have overdone it, convoluting the language with features people love in other languages. But Iā€™m compelled to master them nonetheless.
  • Actually Doing Deep Learning. Deep Learning is really hard to get right, and I care very much about how people manage to do it. My focus here wonā€™t be on theory, but on DL as a craft. (Or as alchemy?)
  • GPU Programming. I think performance engineering is fun as hell, but Iā€™m really not very good at it. (ā€œAndrei is too much of a theoreticianā€ or ā€œphilosopherā€ are accusations my colleagues level against me regularly.) Iā€™m compelled to become much better at it.

My interests are actually at the intersection of software engineering and artificial intelligence research. I used to be an academic who worked on organizing Robotics, Computer Vision, Augmented Reality, good old-fashioned AI and Machine Learning using software engineering principles. Then, I did R&D for 11 years at a nearly-century-old Japanese manufacturing company. I didnā€™t work as much in that intersection like I thought I would. So Iā€™ve stopped doing that, with the intention of returning to my root interests.

Posts

subscribe via RSS