(Also, we set a new one-class weekly record for use this week! This is a big Principles class at a large public univ, but 39k uses of Intelligent Quiz (an AI-driven adaptive quiz) + 3k direct questions in one week. Wild. Across term- with exams + your good setup- 40-80 uses/student is normal.)

So much studying & revision in All Day TA classes happens via platform that I can see exact day, time, nature & type of studying students are doing. Gap (in my classes but also for customers) between exam weeks and others makes me thing we need more testing, not less, and more spread out over term.

And in general LLMs are both better coders and a better fit for "open languages". I think you would be blown away using Codex or Claude Code with R or Python - it is just native, not "copy and paste".

Kelly et al is one I love pointing people to. For general reading, Editor's Introduction at PhD level, Industrious Land at lower levels. For anyone, Lever of Riches. For people of a theoretical bent, Gifts. For Irish-Americans named Kevin who wonder why the Famine barely appears, Why Ireland ;-).

PS - 1) Yes, we're building this out even further. Everything a student does, it should be as close as possible to a tutor who knows what you want to teach sitting by their side. 2) And All Day TA also *reports back to you* summaries of these conversations so you know where students went wrong!

The path forward for higher ed w/ AI is that teaching *complements* AI. Students learn more using AI, and learn *exactly* what we want to teach, instead of just cheating. Can do even better than pre-AI! (And if you want this - alldayta.com, $100 per term. Super easy.)

For students, *I don't care* if they get right answers. Why? If wrong, they have to explain to AI what they were thinking before moving on. Cheating doesn't save time b/c I don't grade on correctness, just whether you work through the quiz! Try it here app.alldayta.com/university-o... 3/4

Toss my lecture audio, slides, handouts into a module. A crazy AI workflow pulls out learning goals. AI then spins up questions (you can approve or not) from your docs plus context on why students might get them wrong. 2/4

Wasn't sure this particular one was a good fit for the bsky crowd, hah!

Hah! Yeah, no one really studies "social science", but eg at an engineering crowd talk, I wouldn't bat at eye at an economist or a political scientist saying "as a social scientist, how we should look at this is..." or similar

A much more common term in the US than Europe. Social Science departments and even high school classes called "social science" are very common over here so that self description wouldn't make anyone bat an eye.

Syllabus here: kevinbryanecon.com/Bryan-Progre...

Course AI here on AllDayTA (to be updated weekly as we progress, including with AI-driven adaptive quizzes!): app.alldayta.com/university-o... 2/2

I cover project setup, version control, my daily very simple workflow, what to use for code (Python or R), what AI is high value, why LaTeX, how to do it easily, why all this matters even for qual projects, and links to exactly what to d/l. My own practices were very sloppy-this is a better way. 2/2

PS - An awesome dev of ours was testing the featureand told me "I got it wrong on purpose at first for testing, but then forgot to divide by 2 for expected value until the system brought me there!" Exactly. Imagine this help for the student, and then summed up & reported back to you for each hw!

Honestly, it's a really nice system. As always, everything is all siloed: your content is never used for any training nor leaves your course, and is deleted on demand. The whole alldayta.com is $100 per class section per term - a couple bucks per student on average. AI complements our teaching! 6/6