BasicAI AGI Seminar


BasicAI is running a free online seminar on artificial general intelligence. The goal is both to bring newcomers up to speed on AGI and to provide a place for researchers working on AGI to exchange ideas.

We will concentrate on technical material relevant to AGI (including various topics from AI, cognitive science, and linguistics), as well as the big picture of AGI: what it requires, why it's such a tricky problem, etc.

Future meetings

Meetings will be held Sundays at 18:00 UTC. The next will be 10/22/17, when we will discuss cognitive architectures (such as ACT-R and Soar). The duration is around an hour.

We are using Zoom, meeting ID number 176-370-982.

To receive announcements of future seminars, add yourself to our mailing list or follow @BasicAIResearch on Twitter.

The current plan is to record meetings to accomodate people who weren't able to make it. Note that recordings may not be available indefinitely.

Past meetings

Our first meeting was held 9/24/17. We discussed AGI in general as well as some of the work done by the "Fluid Analogies Research Group" headed by Doug Hofstadter. Slides here. Note there are lulls in the video when people were chatting via text; chat transcript here. You may also be interested in this blog post.

Target audience

(1) Anyone interested in getting into AGI research (or already involved in it). (2) Anyone who wants a deeper technical understanding of the problem.

All educational/career levels are welcome, from advanced/ambitious high school students to professional researchers. The initial sessions will start off assuming no knowledge of AI (only that you know how to program and understand what a mathematical proof is).

Likely future topics

In keeping with BasicAI's research philosophy and the theme of artificial general intelligence, our focus will be on reasoning, language, knowledge, and learning—as exhibited both by computers and by humans. Topics to be addressed include:

  • Overview of AGI
  • Cognitive architecture
  • Automated reasoning (i.e. with formal logic)
  • Human reasoning
  • Semantics
  • Symbol grounding
  • Induction and learning from experience