things I'm watching as a learning technologies and design leader

I had an impromptu chat with a dear friend who I respect for his thoughtfulness and approach to the work1. After the initial “boy, is the nature of our work ever different from back in the olden days when we actually got to build stuff!”, and “it seems like the bulk of our work now is just profoundly unbloggable for all kinds of reasons…”, we eventually got to “so… what kinds of things are you paying attention to?” That’s a big question, worth articulating and sharing.

In August of 20252, anyway. these are the things I’m watching/trying/exploring. Not necessarily adopting or supporting yet, but taking a closer look at.

Practice - Pedagogies/Approaches/Ways-of-working

Tech - Platforms/Applications/Tools

Content authoring

  • Brightspace Creator+ - we’re evaluating it for possible addition to our campus license. It now includes the full H5P, as well as some AI tools (that we can’t evaluate for reasons)
  • H5P - built into UCalgaryBlogs - we’re looking at ramping that up for use across campus (especially if the Creator+ evaluation winds up with us not adopting that tool - but, interesting, since D2L acquired the H5P organization…)

Team Formation and Peer Feedback

This is something of a gap in our tools. We have previously built our own Peer Feedback Tool, but that’s long gone.

  • ITP Metrics - We have an excellent research lab on campus, and I would love to find a way to bake their software into Brightspace so it could be made available in all courses.
  • Feedback Fruits - an interesting peer feedback platform we’re looking at (and waiting for institutional machinery to catch up before being able to do anything with it)

Classroom engagement

Not a great name for this category. Student Response Systems. Things that let you ask questions during a class/workshop/presentation. We have Top Hat as a campus platform, but sometimes need a tool with more flexibility (and ability to use without requiring a login). Other tools we’re looking at include:

  • Mentimeter - licensed at a small scale, for use in TI programming
  • WooClap - dumb, dumb name. Interesting software - we haven’t licensed it, but it’s something to watch.

AI

Microsoft Copilot4 is our officially-approved campus AI platform, but it is… rudimentary and limited. Other AI tools I’m looking at include:

  • AI Applications that I actually use in practice

    • Claude - I’ve personally paid for a Claude Pro license for almost a year. It’s been extremely useful for tasks that benefit from RAG, and the text it creates is surprisingly effective. Especially, using it to convert pages of a paper notebook, whiteboards, post-its into markdown-formatted text. Amazing. I don’t use the Claude Code thing (yet?), but I use Claude a lot for summarizing and I have been seriously impressed by the Research mode.
    • ChatGPT - limited use because OpenAI is problematic, and the image generation is a distraction. I mostly use ChatGPT to make goofy images to share.
  • AI for Teaching & Learning

  • Hosted AI Platforms to build stuff on

    • Azure AI and AI Foundry - possibly as a platform to build AI applications on, as part of our campus MS environment. It’s marketed as an easy AI development/deployment platform, but holy does it look like a complex beast that has the added benefits of further entrenching one vendor while also being rather expensive…
    • AnythingLLM - Docker containers for hosting any LLM. Might scale to something like a mini-campus AI service without needing Azure/AWS etc on the back end? If I had access to some server capacity, I’d try deploying this to see what might be possible.
  • Local LLM platforms

    • Ollama - for running “open” LLMs locally on my macbook air.
    • Ollamac - an app that provides an interface on top of Ollama’s server. This is how I usually play around with local LLMs5.
    • Msty - an all-in-one local LLM interface that can do some RAG. I’ve played with the RAG, but honestly it’s just too slow to use. The RAG in Claude is outstanding, and this… isn’t.
    • Sidekick - an all-in-one local LLM interface that can do some RAG - looks interesting, but I haven’t tried it yet.

I think some of those local LLM tools can also take advantage of higher-end hardware like insane graphics cards and AI accelerators.

Anyway. Some of the stuff I’m tracking at the moment. YMMV. IANAL. etc.


  1. It was a great conversation, which also reminds me that I need to make the effort to actually talk to my mostly-distributed-across-time-and-space friends. (and I’m intentionally not naming the friend - you know who you are.) ↩︎

  2. holy hell. how is it AUGUST of 2025 already??? ↩︎

  3. the conference website hasn’t been updated to reflect the 2026 theme yet… ↩︎

  4. The free/public version, but linked to campus SSO and with extra data protections. Not to be confused with MS365 Copilot (the one that integrates with MS365 apps and can access your files/data/calendars/emails but costs ~$40/person/month), which is different than MS Azure AI and MS AI Foundry. They’re building AI into everything, both as separate products and infused into everything. It’s a mess. ↩︎

  5. I’ve tried running phi3, deepseek, mistral, llama, gemma, etc… - but the storage on my macbook air is not huge so I have to keep an eye on that, and they still run kind of slow-ish on an M2 CPU/GPU… ↩︎

Last updated: August 1, 2025

Comments powered by Talkyard.