Skip to content
View etsabary's full-sized avatar

Block or report etsabary

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
etsabary/README.md

Eldad Tsabary

I'm a professor of music and electroacoustics at Concordia University, a researcher and developer, and founder of the Concordia Laptop Orchestra (CLOrk). My work connects musical creation and collective improvisation with questions about how people listen and create together.

My current research program, The AI Mirror, uses building as a way of investigating human nature. I develop and test generative systems to examine what their behaviour reveals about human creativity and social life, including the assumptions we put into them. I also explore how people can retain agency over what algorithms show them.

One research instrument is The Village, a persistent virtual world whose characters accumulate memories, relationships and ongoing activities. Following their interactions lets me examine what we recognise as human, and question why something seems convincing.

I also publish independent AI benchmarks, comparing decision accuracy alongside response time and cost. These help developers judge which models fit a particular task and where their limitations matter.

Selected work

Pinned Loading

  1. system-1-decision-benchmark system-1-decision-benchmark Public

    System 1 Decision Benchmark v2: 1,934 questions across 41 categories. 26 configurations meet the 62% cutoff: 20 hosted decision models, 3 local models and 3 reasoning-disabled LLM references. Accur…

    HTML 4

  2. visual-decision-benchmark visual-decision-benchmark Public

    200 image-grounded decisions: 21 complete routes across 20 categories, plus 20-question speed screens including two Apple M2 models. Accuracy, latency, cost and category profiles.

    HTML 1