What might new developments in artificial intelligence mean for higher education?
I’ve been exploring this topic for years, and now would like to announce a new project about it. I’m launching a Substack newsletter, AI and Academia, with the goal of focusing my AI work in one spot. Plus I get to try out Substack.
Here’s the first, introductory post.
So why take this particular step? Why use Substack, which has various issues?
Several reasons. First, I’d like to have a dedicated spot for my AI and education research, and this seems to fit that need neatly. The subject is deep enough and potentially important enough to have its own home, rather than being a strand within a larger site. (I have been working on a resource website, too. Will announce when it is decent, unless something makes it redundant.)
Second, I’m curious about how Substack newsletter facilitate conversations. Those of you who know my work know that I’m focused on getting discussions going. How will readers use this email and web-posted transmissions as prompts? Will the Notes function actually work? I hope to add this platform to my decades-long exploration of online communication.
Third, I’d like to see if the Substack economic model will help sustain my work on the topic. Researching AI and forecasting its academic implications is a major enterprise, and one I don’t currently have any support for, as an independent. Learning desktop AI packages, keeping up with an explosively fast field, tracking instances and impacts across global higher education, and more – this takes a significant amount of time plus computational costs. Maybe the Substack model can support me doing the work. I’m tinkering with it as an experiment.
I’ll continue sharing AI research in other venues, like the Future Trends Forum (three sessions are in the pipeline now!), presentations, workshops, and classes. Depending on how things go, I might do more.
In the meantime, check out AI and Academia and see what you think!
(photo taken by human being Alan Levine way back in 2007; titles automatically generated by Substack software, I think)
Good idea, Bryan. On your resource list, put Weizenberg first. He not only explains the software, he shows what happens when humans fail the Turing test — they can be convinced of anything as long as their conversational partner can come up with the right next word. An absolute must signpost.
And maybe some of Elizabeth Loftus’s work, showing that presuppositions can’t be distinguished from memories internally — there is no such thing as eyewitness memory.