Agreed. #CAI is doing two 'boot camps' for our campus and will be sharing the accompanying (NEH-funded) starter kit more widely.

The boot camp team is also available for remote or on-campus presentations

Anna Mills, annamillsoer.bsky.social, she/her Critical AI : first issue out! https://read.dukeup Sal Khan outlines exactly this in his book about his AI tutor in general, and gets very specific about it in this recent talk, real time monitoring of student writing against a rubric, telling the student what grade to expect. It's about 6h38m in: youtube.com/watch?v=wVqNLa…

More Me-lon Usk disinformation.

🎧 'What is not acceptable is when my art and my identity can simply be taken by a third party and exploited falsely for their own gain.' - #SagAftraMember FKA Twigs

Check out the newest SAG-AFTRA podcast episode on A.I. developments. Listen now: ow.ly/tIXT50Sc8hr

Thinking of this kind is crucial IMO to #criticalailiteracies .

Teaching 'harms' only to move on to convenient use cases is arguably hypocrisy

Square the circle by teaching students to understand bots as researchers, not consumers.
eripsa (professional robot) and Jane Rosenzweig do this well!

Important food for thought from Safiya Umoja Noble & Sarah T. Roberts. This point is highly relevant to conversations that have been happening on here about GAI / LLMs in education--not to mention Ed Tech more broadly.

Good reporting from @nyt's Kashmir Hill + Tiffany Hsu-in contrast to the mixed bag that Thomas Edsall published last week which reproduced a great deal of unquestioned hype and junk science and 'reported' it alongside purely speculative thinking abt AI. Some snips:

nytimes.com/2024/06/06/tec…

None of us has read this yet but we LOVE when students publish their work on AI related subjects! Congrats to at Jane and of course to her student!

Ready for more Mystery AI Hype Theater 3000?

Join me & Alex Hanna (اليكس حنٌا) on June 10, when we get to chat with Prof. Safiya Umoja Noble about Google's AI Overviews -- in the long context of Google's warping of information access.

Monday June 10, noon Pacific
twitch.tv/dair_institute

Aravind Srinivas Perplexity The screenshot you shared that has prominent sourcing at the top is not what I'm talking about. This page, which reads like an original news article, does not have appropriate attribution. perplexity.ai/page/Eric-Schm…

I heard someone say recently that Perplexity was better than Google's AI Overviews and OpenAI because it 'cited its sources.'

Unfortunately LLMs do not know what a claim is, and cannot actually 'cite' anything. Especially as they are trained on stolen data as it is.