Doug Holtondougholton
2025-10-19
Doug Holton boosted:
2025-10-12
Doug Holtondougholton
2025-10-10

Cool demo by @TodePond of a classroom agent in @tldraw
youtu.be/L-fKtZXDb-s?si=OEoYVR
Jump to the 10 minute mark to see an example of AI giving guidance to a student working on a math problem on the tldraw digital whiteboard/canvas. Uses the agent starter kit tldraw.dev/starter-kits/agent

Doug Holtondougholton
2025-10-08

@rhettallain yeah I tried getting ChatGPT and Gemini to generate graphs that correspond to calculus problems and it was pretty bad. I think code generation, Python plotting might work better, but haven't tried it, and that's not so accessible for students

Doug Holton boosted:
Jason Neiffer, Ed.D.neif@mastodon.cloud
2025-10-05

This framing gets it. Assessment in the AI era isn’t a puzzle to solve; it’s a balancing act we’ll never finish. We should expect messiness and make space for it institutionally.

onedtech.philhillaa.com/p/on-f

Doug Holton boosted:
2025-10-03

A principled way to think about AI in education: guidance for action based on goals, models of human learning, and use of technologies

Noah D. Finkelstein
arxiv.org/abs/2510.01467 arxiv.org/pdf/2510.01467 arxiv.org/html/2510.01467

arXiv:2510.01467v1 Announce Type: new
Abstract: The rapid emergence of generative artificial intelligence (AI) and related technologies has the potential to dramatically influence higher education, raising questions about the roles of institutions, educators, and students in a technology-rich future. While existing discourse often emphasizes either the promise and peril of AI or its immediate implementation, this paper advances a third path: a principled framework for guiding the use of AI in teaching and learning. Drawing on decades of scholarship in the learning sciences and uses of technology in education, I articulate a set of principles that connect broad our educational goalsto actionable practices. These principles clarify the respective roles of educators, learners, and technologies in shaping curricula, designing instruction, assessing learning, and cultivating community. The piece illustrates how a principled approach enables higher education to harness new tools while preserving its fundamental mission: advancing meaningful learning, supporting democratic societies, and preparing students for dynamic futures. Ultimately, this framework seeks to ensure that AI augments rather than displaces human capacities, aligning technology use with enduring educational values and goals.

toXiv_bot_toot

Doug Holtondougholton
2025-10-02

A couple of sample notebooks related to :
* The Nature and Mitigation of AI Sycophancy notebooklm.google.com/notebook including a note (under Studio tab) on implications for college
* College Success (OpenStax textbook) notebooklm.google.com/notebook
You could upload your textbook (if ) or course notes and share the link with your students to make a course-specific AI resource. While not required, Gemini Pro is free for a year for students gemini.google/students/

Doug Holton boosted:
2025-10-02

It should be obvious a priori that it's absurd to judge colleges by the number of applicants they manage to avoid admitting. Quality ≠ selectivity! Now there's evidence demonstrating the absence of an empirical relationship between the degree of intellectual engagement at a school and that school's selectivity: nytimes.com/2025/09/22/opinion

So what does actual educational quality look like? For decades, researchers at Indiana University have been measuring something that matters far more than acceptance rates: how engaged students actually are in their learning. Through the National Survey of Student Engagement, they survey hundreds of thousands of freshmen and seniors about how they spend their time, interact with professors and what they think they are gaining from college.

While the survey results are not routinely made public, I asked Indiana researchers to compare data across groups of schools sorted into five tiers based on selectivity. Here's what they found: On 18 different engagement measures, the tiers of schools were separated by just a point or two among freshmen — a gap that narrowed even more by senior year.

For instance, in a measure of “higher-order learning” — engagement in activities that promote the critical, analytical thinking central to a college education and essential to jobs in an age of A.I — there was “no statistically significant difference” between the most selective schools (under 20 percent acceptance rate) and the next tier down (20 percent to 40 percent). Overall student satisfaction among seniors was separated by only four percentage points between the most selective and least selective schools (those with an 80 percent to 100 percent acceptance rate).
Doug Holtondougholton
2025-10-02

Notes and system instructions and deep research reports on creating a custom GPT or Gemini Gem to tutor calculus: docs.google.com/document/d/1p2
Hawkulus: AI Calculus Tutor (Custom GPT): chatgpt.com/g/g-68d992b987fc81
Newton
Newton: Calculus Learning Coach (Gemini Gem): gemini.google.com/gem/1IojGkCE

Doug Holtondougholton
2025-10-01

Asset-based Approaches to Student Success
edtechdev.wordpress.com/2025/1
In the past, I’ve shared strategies that have been shown to improve , but none of those are effective if we or our students have a fixed mindset (belief that intelligence is fixed) or deficit thinking (focusing on what students lack instead of what they bring to the table). The alternatives are growth mindset (intelligence can grow with effort) and asset-based approaches).

Doug Holtondougholton
2025-09-26

Teaching Tips: The First Major Exam or Assignment
edtechdev.wordpress.com/2025/0
A sampling of some research-based practices you may find useful for before, during, and after your first major assessment to improve student success and help struggling students turn things around afterward.

Doug Holton boosted:
Human G-G-G-GhostwriterHG@beige.party
2025-09-25

A friend of mine is a professor who teaches CS and bioinformatics classes. She uploaded the wrong exam for her students, and a huge portion of them (like, 85% of the class) nailed material they had never seen before. Like, 100% quality work on Python or whatever (computers are just magic to me). And then one of the students asked what the deal was because they had never studied this stuff, and she got to learn that nearly all of her class was GPTing their way through it.

Doug Holton boosted:
2025-09-25

This looks interesting

LAMB - Learning Assistants Manager and Builder

"LAMB (Learning Assistants Manager and Builder) is an open-source web platform that enables educators to design, test, and publish AI-based learning assistants into your Learning Management System (LMS like #Moodle without writing any code."

github.com/Lamb-Project/lamb

#edtech #teaching #ai #chatgpt

Doug Holtondougholton
2025-09-16

Recent articles:
* Collaborative Learning Activities For Promoting AI Literacy arxiv.org/abs/2508.15111v1
* Analyzing Problem-Solving in Physics Through Interaction With an AI Chatbot arxiv.org/abs/2508.14778v1
* MathBuddy: A Multimodal System for Affective Math Tutoring arxiv.org/abs/2508.19993v1 code: github.com/CleoChu/MathBuddy
* How Math Teachers Are Making Decisions About Using AI hai.stanford.edu/news/how-math
* How does AI affect how we learn? thecognitivepsychologist.subst

Doug Holton boosted:
The Conversation U.S.TheConversationUS@newsie.social
2025-09-07

The myth of scientific objectivity hides the truth that every experiment is influenced by cultural values, assumptions and beliefs – and that actually hurts the search for truth

Analysis:
theconversation.com/scientific

Doug Holtondougholton
2025-09-05

Using AI to create unlimited practice problems [in statistics] youtube.com/watch?v=tsmHmjBfGwI
chatgpt.com/g/g-67d1da836ac081 More on creating custom GPTs to generate problems edunewsletter.openai.com/p/tea
They've also developed a custom tool that lets instructors see data on the way the students use it. Looking for instructors to test: pingpong.hks.harvard.edu/eduac
For a related open source project, see OnMicro for developing educational AI apps onmicro.ai/ github.com/onmicroai/micro_ai/

Doug Holtondougholton
2025-09-04

Workado promoted its AI Content Detector as 98% accurate in detecting whether text was written by AI or a human. However, independent testing showed that the accuracy rate for general-purpose content was only 53%. The FTC alleges that Workado violated the FTC Act because the 98% claim was "false, misleading, or non-substantiated." nitter.net/LuizaJarovsky/statu
More on the numerous problems and issues with AI detectors: dangerousaideas.substack.com/p

Doug Holton boosted:
Curated Hacker NewsCuratedHackerNews
2025-09-04

Evidence that AI is destroying jobs for young people

derekthompson.org/p/the-eviden

Doug Holtondougholton
2025-08-25

Recent articles:
* A Comprehensive Review of AI-based Intelligent Tutoring Systems arxiv.org/abs/2507.18882v1
* AI-Powered Math Tutoring: Platform for Personalized & Adaptive Education arxiv.org/abs/2507.12484v1
* Automatic LLM Creation of Interactive Learning Lessons arxiv.org/abs/2506.17356v1
* LearnLens: LLM-Enabled Personalised, Curriculum-Grounded Feedback with Educators in the Loop arxiv.org/abs/2507.04295v3
* AI Playbook for Teaching & Learning Leaders books.lib.uoguelph.ca/aiplaybo

Client Info

Server: https://mastodon.social
Version: 2025.07
Repository: https://github.com/cyevgeniy/lmst