I’ve been using, developing, implementing, supporting, and critiquing educational technologies for longer than I care to admit. In that time, roughly once every decade, a major new technology emerged that was simultaneously cast as both the future and death of education. Computers in the 1980s. Multimedia and the Internet in the 1990s. “Web 2.0” in the 2000s. MOOCs in the 2010s. And now, generative AI (GenAI).
I don’t look at the webserver logs for this webthing in detail, but have been curious given the spikes in bot traffic over the last several months. Much of it is the traditional “I’ll just crawl your entire website multiple times per day to feed my search index”, but there’s been a LOT of “I’ll just crawl your entire website multiple times per day to feed my LLM training dataset, and I will ignore rate limiting that you’ve set in robots.txt”
Deja vu all over again. I did this the last the the bad orange man was elected, and before that back in 2008, but this time I’m going a bit further. It sucks. This is hard, and I know it looks performative and theatrical and melodramatic, but it’s something that I think I have to do. I’m sad and frustrated and pissed off that I even have to think about this nonsense. I’m conflicted on so many levels.
I’ve been using Claude since the 3.5 model was released, and have been paying for the Professional tier for the last 5 months (and counting). $30 CAD per month for a tool that I use pretty regularly. All AI companies are problematic, but Anthropic feels slightly-less-problematic than OpenAI, certainly less so than Grok. Claude feels less gimmicky - it is strictly a text generator and doesn’t generate images. Its ability to build simple software is pretty impressive, even if the process can be mind-numbingly frustrating at times.
I’ve been using Obsidian as my note-taking app for a couple of years now. I’ve got over 5,000 notes for various things. Including lots of meetings. There are times when my job is basically going to meetings. And I need to be able to keep track of what we intend to talk about in each meeting (especially as they start to blur together).
Graph of over 5,000 notes in Obsidian. The red dots are mostly meetings. Blue are people, Orange-ish are Topics. Green are vendors.
It’s like the 2007 blogosphere may be returning1? Hopefully. Anyway, Alan tagged me in this chain letter blog challenge thing, so here goes…
Why did you start blogging in the first place?
My first blog posts were written in the then-new Apple iTools “iWeb Publish” tool. I have vague memories of static pages being saved to my iDisk, then synced to a webspace on an Apple server somewhere. I’m not sure why I started posting on that. Just to try it out? Maybe there’s something to this whole Web thing? I’d started just before going to my first WWDC, so was drinking the kool-aid pretty deeply at the time. Early posts were about our not-yet-born child (who is 22 years old as I write this), WWDC, and learning objects software development stuff from the early CAREO days.
This fall marked 30 years that I’ve been working in edtech. I’d done some edtech projects before then, but Fall 1994 was when I started doing it for a living. 30 years is a long time, simultaneously forever and gone in a flash. Instead of writing a 50,000 word series of posts documenting minutae, I took some time to reflect on some of the major themes and changes that have defined my career so far. (the minutuae are already documented on my Projects page, in my Ancient archive, and in my CV.)
The growing use of GenAI in education, technology, and society raises four main conflicts: misunderstanding its capabilities, insatiable resource hunger, outsourcing creativity, and complicity in a problematic system.
I finally tried out Google’s newly-announced NotebookLM generative AI application. It provides a set of LLM-powered tools to summarize documents. I fed it my dissertation, and am surprised at how useful the output would be.
The most impressive tool creates a podcast episode, complete with dual hosts in conversation about the document. First - these are AI-generated hosts. Synthetic voices, speaking for synthetic hosts. And holy moly is it effective. Second - although I’d initially thought the conversational summary would be a dumb gimmick, it is surprisingly powerful.
I explored Pavel Samsonov’s article on the limitations of being ‘data driven,’ reflecting on the importance of being ‘data informed’ in complex work environments where trust and collaboration are essential.
Hugo is the static site generating content management system that I use to publish this website. It works really well, and has some deep functionality that I’m not even touching. For instance, it can parse data files while generating the site - including JSON and XML - and can use the content of those files to display information on web pages.
I was going to follow somerecipes that I found online, but they involved converting the OPML file into JSON to be read by Hugo. I didn’t want to do that if possible. So, time to roll my own solution using built-in functionality in Hugo…
In recent years, insurance companies have offered incentives to people who install dongles in their cars or download smartphone apps that monitor their driving, including how much they drive, how fast they take corners, how hard they hit the brakes and whether they speed. But “drivers are historically reluctant to participate in these programs,” as Ford Motor put it in a patent application that describes what is happening instead: Car companies are collecting information directly from internet-connected vehicles for use by the insurance industry.
Graph of over 5,000 notes in Obsidian. The red dots are mostly meetings. Blue are people, Orange-ish are Topics. Green are vendors.