With the release of GPT-4o, we're yet again arguing about the capacities of LLMs, and how to conceptualize their agency. In a #FAccT2024 paper, I argue we have a very clear framework for understanding ChatGPT and similar interactive AI systems: as animated characters
1/n

Just when you think you've seen everything. 😂

And yet more fantastic news to share. Waterloo's team of algorithm programmers won 18th overall at the 46th ICPC World Finals! #UWaterloo

Congrats, Marian Dietz ,Chris Trevisan and Wen Yuen Pang! 👏

scoreboard.icpc.global/46/index.html

narcissistic Canada Goose

Please submit your papers to our ARIAL Workshop to be held with IJCAIconf , Jeju, South Korea. Accepted papers will be submitted to Algorithms MDPI special issue on AI and Healthcare.
Reza Basiri Ali Abedi

The good news keeps coming in ... CS Prof Yaoliang Yu, a faculty member at the Vector Institute and a Canada CIFAR AI Chair, has received an Ontario Early Researcher Award to develop pushing-forward deep generative models. Congrats, Yaoliang!

cs.uwaterloo.ca/news/yaoliang-…

Congratulations to CS Professor Timothy Brecht, recipient of a ~$60k Alliance grant for his project “Real-Time Hockey-Analytics Using Puck and Player Tracking Data”!

Congratulations to CS Prof Ian Goldberg, recipient of $800,000 through the Innovation Fund under the Canada Foundation for Innovation for his project “UPSCOPE: Understanding Privacy, Security, and Cryptography in Online and Physical Environments.”

uwaterloo.ca/news/waterloo-…

Congratulations to UWaterloo English's Dr. Heather Love, whose recently published monograph, Cybernetic Aesthetics: Modernist Networks of Information and Data has been shortlisted for the 2023 British Society for Literature and Science Book Prize! tinyurl.com/bdesduf5

Thrilled to announce the One Million Deepfakes Detection Challenge at ACM Multimedia 2024!

Please consider participating! Challenge early page can be accessed at deepfakes1m.github.io

with Shreya Ghosh Zhixi Cai

The problem is not what kind of car you drive (EV, gas, hydrogen, Mr. Fusion, self-driving, flying...). The problem is that you are not riding a bicycle.

Respectfully, AI today is not like atomic science before the Manhattan project. The problem is there's a terrible Wizard of Oz effect. Even smart people who understand LLMs intellectually get fooled by the seeming intelligence of ChatGPT etc. Smooth text interpolation + a giant

My latest paper draws a parallel between Sartre's 'Being and Nothingness' and the free energy principle. osf.io/p9z2j via Center for Open Science

.
UWaterloo English's Dr. Heather Love will be giving a talk, 'Cybernetic Aesthetics,' as part of the WICI Speaker Series on Mon. Jan. 29, 1-2:30pm, DC 1302. tinyurl.com/4ud4uv42