Below is another set of IJCAI session notes. This was the first invited talk in a day-long workshop called Architectures and Evaluation for Generality, Autonomy & Progress in AI (AEGAP). The speaker, Oren Etzioni, talked about some of the work the Allen Institute is doing to drive the creation of common sense in AI. He focused especially on a need for a concrete benchmark to measure results when it comes to implementing common sense. As before, these are mostly for myself, to sort through the notes in my head and revisit something a little more organized in the future. In addition to the talk itself these will cover the Q&A which followed.

For the past couple of days I’ve been at IJCAI - the International Joint Conference on Artificial Intelligence. It’s been a three-day whirlwind of workshops and tutorials. I took a lot of notes, and as with ECAL last year I figured organizing them into blog posts would be a good way to review some of the takeaways. Note that the intended audience here is future-me, and I could very well have made mistakes or misinterpretations in my own notes. Also apologies about any typos.

I should know better than to use a latest tag in a Docker image. I’d never do this at work, so why did I do it in my hobby project? I blame Go’s nature of dealing with package versioning (that is, the lack thereof) for putting me in the “let’s use latest!” mindset (and yes I know they’re working on it). Anyway, I should know better, but here we are, writing this post. It’s going to be a short one.

This post will have some general impressions of Input Club K-Type and matt3o’s new keycap profile (MT3), with the /dev/tty color scheme. Both products were ordered from Massdrop last year.

I’ve started working on client tests, which have been getting very neglected compared to the server. I figured I’d write some quick notes on what I’ve done so far before continuing, before I forget.