As the new semester begins, there's been a big uptick in views of my "Four Rules for Writing in the Age of AI." I'm working on a new set of rules, but in the meantime, I wanted to share the originals again. Link in bio since if I put it here you'll never see this post. /1 But also, short version of the rules here: Rule #1: Understand how LLMs work. Teachers, students, admins, all of us need to know how this tech works to make informed decisions about whether/how to use it. Start here: /2understandingai.org/p/large-langua…

There will be 10K articles on what OpenAI's new features "mean" for education by the end of the day. Some cool stuff for sure--real time translation; vision. But here's what's still true: 1) getting real time audio answers to every question could enhance learning /1 or it could remove important friction and sap motivation; 2) students will need to have skills and know how to do their own thinking to use any of this effectively; 3)students are already struggling to find meaning in learning and these tools may or may not help/hurt; /2

If I were reading this post with my students, I would ask them why they think someone might find it surprising that people would find these ideas controversial. I'd ask, what assumptions might someone have that would make them think everyone would agree with these statments/1 Here's what would likely emerge from that discussion (based on years of reading closely with students): They would identify these assumptions in the first post: 1)the technology that is currently being discussed will create an abundance of intelligence, energy, longevity, etc./2

Just this weekend I've read about AI-generated fake literary agents, AI-generated fake books by real authors that are now part of their Goodreads profiles, AI-generated fake travel guides. Starting a thread of threads to keep track of these writing-related AI impersonations. First up, fake travel books written by people who are supposedly "renowned travel writers" (one is "Mike Steves" perhaps to confuse people looking for books by Rick Steves." nytimes.com/2023/08/05/tra…

A depressing read. washingtonpost.com/technology/202… We learn that 1) People are already losing their writing jobs to ChatGPT, even though it doesn't do a great job: "for many companies, the cost-cutting is worth a drop in quality." 2)The examples in this article challenge the narrative that ChatGPT will make us all more productive and make all our work better; it's just as likely to make the work worse. This, @ubiquity75:“We have to ask: Is a facsimile good enough? Is imitation good enough?