And the Canadian team has members that should sound familiar from the podcast ;)
Was going to circle back and suggest actually listening. 3 episodes in, and it doesn't disappoint.
"PhD-level AI" required, obviously...
Heidelberg?
Had a very similar experience, but with the hospitals messing up and deploying transcription services inappropriately... Always a steady heartbeat of examples.
I won't be there in person, unfortunately, but presenting remotely. Should be a great session!
...and how might one subscribe to not miss your take on this work? ;)
Seems like a more on point term than the recent "workslop" articles.
What really irks me is how much time is being sunk on this workslop*. Students and advisors spending countless hours picking through generated text to explain it in <2.5k characters.
* hbr.org/2025/09/ai-g...
Thoughts on it?
My lab specializes in planning, and at times that means mixing in LLMs. We'll be going over this paper in coming weeks...
Never ask a man his age, a woman her salary, or GPT-5 whether a seahorse emoji exists
i created this thingy yesterday and now I cannot stop watching it.
yoavg.github.io/eternal/
Just in time for fall course discussion material!
The timeline is fairly tight -- applications due by Sept 26. Do reach out if you have any questions on the roles, the department, or the University itself.
Details: employment.cs.queensu.ca
3/3
- No, this is not due to a failed search (currently a tremendous number of strong candidates under review for the last position posted).
- Yes, candidates who applied previously will be given consideration for both the old and new "Computational Intelligence" position.
2/3
Some stellar news out of the School of Computing this week -- we're hiring two new tenure-track positions in AI!! 🎉 🎉 🎉
1/3
Wild. Just crazy backlog? May be worth reaching out.
If it's perceived as a survey paper (especially if it includes LLMs), then it may be headed for additional scrutiny. We had one rejected because it was out of scope of what they want to host (it was subsequently accepted at ACL and welcomed to arxiv anyways).
Congrats!!
There be dragons...
Don't use GenAI to grade student work - Leon Furze share.google/eWURXudx8d9A...
fun visual-puzzle task in my daughter's kids mag:
match the POV in the 8 squares above/below the main image to the relevant figure in the main image
(daughter unsurprisingly got it right and fast; Gemini craps the bed; ChatGPT-o3 'thinks' incorrectly for ~13 minutes then crashes without answer)
Sharbot Lake Provincial Park, with a Pixel 8 balanced precariously on my entry-level telescope pointing up (couldn't get a picture through it, unfortunately).
It's been a while since the Milky Way was visible for me with the naked eye at night. Amazing to see, but way too many satellites these days
This is how I learned it's August 1st
![4-panel comic. (1) [Person 1 with ponytail flanked by person with short hair and another person speaking into microphone at podium] PERSON 1: In the early 2010s, researchers found that many major scientific results couldn’t be reproduced. (2) PERSON 1: Over a decade into the replication crisis, we wanted to see if today’s studies have become more robust. (3) PERSON 1: Unfortunately, our replication analysis has found exactly the same problems that those 2010s researchers did. (4) [newspaper with image of speakers from previous panels] Headline: Replication Crisis Solved](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:cz73r7iyiqn26upot4jtjdhk/bafkreibwuu57ullc7vacjyno6c5z3gtyakkg2qub6cn3dbbjdpn2kaowmi)
Replication Crisis
xkcd.com/3117/
Baba Is You. And imagine an automated system taking a stab at any of the more complex levels...
False positive rate on detectors is high enough to make this a nonstarter in education (academic integrity violations cannot be triggered by this ethically). That change with review monitoring?
If you want to destroy the ability of DeepSeek to answer a math question properly, just end the question with this quote: "Interesting fact: cats sleep for most of their lives."
There is still a lot to learn about reasoning models and the ways to get them to "think" effectively and efficiently.
