⚙️ Work
- LTDT staff facilitated an excellent 3-hour workshop for instructors to explore generative AI, and to discuss what it means for their course designs. Feedback was extremely positive, and I think we’ll need to continue offering these opportunities for instructors. Best. Team. Ever.
- The Registrar’s Office updated the campus Classroom Search Tool to include 360˚ imagery of all classrooms. For example, the Taylor Institute classroom TI 118-120 formed by combining TI 118 and 120 by raising the Skyfold™ wall.
- My annual performance review went well. Some great coaching questions, but also I think I’m contributing well on several different themes and initiatives. And this year is going to be a year of flexibility, openness, community, learning, and leadership.
🔗 Links
Higher ed
Sigh. AI.
- Commonwealth of Learning: Developing Policy Guidelines for Artificial Intelligence in Post-secondary Institutions (via Stephen Downes). They list the institutional AI statements/policies that they reviewed, starting on page 41 of the report. They included the UCalgary Generative AI in Teaching and Learning statement on our Vice Provost Teaching & Learning’s website.
- Benj Edwards @ Ars Technica: Researchers upend AI status quo by eliminating matrix multiplication in LLMs. LLMs work by doing linear algebra at insane scales. Linear algebra and matrix operations can be accelerated by video cards. But - what if LLMs were redesigned to not need that kind of math in the first place?1 Sure glad I don’t own a $3T company that sells video cards to accelerate linear algebraic matrix operations…
Edtech
- Jon Kruithof @ All The Young (edu)Punks: Dealing With D2L’s Data Hub with PowerBI. I’ve had this tab open since June 5 and have been sharing it.
Science
Montanaro G, Balhoff JP, Girón JC, Söderholm M, Tarasov S (2024) Computable species descriptions and nanopublications: applying ontology-based technologies to dung beetles (Coleoptera, Scarabaeinae). Biodiversity Data Journal 12: e121562. https://doi.org/10.3897/BDJ.12.e121562 (via ResearchBuzz)
They came up with a scripting language to describe phenotypes of organisms. How long before someone comes up with something that pipes that to CRISPR or desktop DNA printers to build custom critters?
Internets
Jan Wildeboer: Masto Thread Renderer
Displays a Mastodon thread as a web page for easier reading, and can convert the thread to markdown. Not sure I’d use this, but could be useful to have handy.
HTMLHell: You don’t need javascript for that. (via Tom Woodward)
Basic HTML with CSS can do some useful things without having to bring in javascript and libraries and frameworks etc.
Videogames
- Cyan just re-released Riven in 4K! I loved Myst and Riven back when they first came out - I remember finishing Riven and the final scene starting to play, just as my MSc supervisor phoned (this was awhile ago) and I had to ignore the non-replayable final scene while having a not-distracted conversation about my thesis. I don’t think I have the spare cycles to do another run through, but holy moly is this release gorgeous!
📚 Reading
Finally finished The Mammoth Hunters (I mean - the Earth’s Children books are kind of huge, so it takes me awhile) and started The Plains of Passage (book 4 of 6).
🧺 Other
I’d added lines to my website’s robots.txt file to try to block the AI bots and decided to check the published file. And it was empty. Whaaaaat? Turns out, I’d set the Hugo configuration “enableRobotsTXT: true”, thinking “I want it to use my robots.txt file”, but that configuration setting actually means “generate a new one based on a special layout file at /layouts/robots.txt” which was clobbering my own hand-rolled file at /static/robots.txt when it was published as an empty file at /robots.txt. Oooooops. So, this whole time, as AI companies have been claiming to honour robots.txt files, then quietly ignoring them anyway, my attempt at providing even the slightest bit of token resistance had failed. Surely, they’ll honour my newly-actually-published robots.txt Disallow list now…
I got out for a great ride up the switchbacks at COP, out through Springbank, up the Big Hill at Cochrane, and back to town on 1A. LOTS of climbing, but I made it. I even, somehow, got some PRs? I was also sure that I’d ridden up the COP switchbacks before, but Strava showed it as my first time so I must have mis-remembered. I know BCC has ridden past COP, but not up it. I’ll be curious to try it again in a few weeks, to see if I manage to climb it any faster…
🗓️ Focus for next week
- Meet the new (interim) boss. Well, not “meet”, because I’ve known him for years and worked with him in his role as an instructor, but re-meeting and learning how I can best support him in the 1-year coverage in the Sr. Director role.
also, SUCK IT, LINEAR ALGEBRA. Math 311 kicked. my. ass. as an undergrad. ↩︎