The Bot Brothers: AI for Educators

5 easy steps for responsibly piloting AI and tech in education gettingsmart.com/2024/08/06/5-e… via @Getting_Smart

@AnnaRMills Yep. I was just thinking about this today. I was also thinking is this the way writing is going, and should I be working with kids on this?

This makes full disclosure of AI's role in a writing process tricky. ChatGPT doesn't give a share option for sessions with its Canvas feature. The AIarchives.org extension does create a link to the side chat part if not the document and its revisions.

@Marc__Watkins It seems to me that besides specific use cases, like teacher coded tutors, there is less and less reason to employ AI in education!

My latest piece for the Chronicle about AI reading assistants: And that’s the most frustrating aspect of generative AI in education: This is technology that can truly help some students yet can also, through misuse or over reliance, seriously weaken the skills of many others.

I asked ChatGPTo to work as a copilot with me in creating a rubric. It shifted to Canvas mode. After creation, I asked for suggestions, it highlighted the text and added commentary. Game changer. #APLang

Definitely worth a read, educators.

@emollick I’m curious as to why this isn’t just a GPT? Is it so that it is LLM agnostic?

We wrote a piece on how to create prompts that create prompts, with a focus on how teachers can create reusable tools for tasks (we call these prompts blueprints) The full text of all the prompts is included, as are GPTs if you just want to try them out. hbsp.harvard.edu/inspiring-mind…

I don’t often look at OpenAI’s site but I did today and came across this page. I think it is helpful for those who teach and worry about writing. Thoughts? openai.com/chatgpt/use-ca…

@emollick Seems like that should be rather obvious but data for backing seems to be a necessary requirement in the 21st century.

AI can help learning... when it isn't a crutch. There are now multiple controlled experiments showing that students who use AI to get answers to problems hurts learning (even though they think they are learning), but that students who use AI as a tutor perform better on tests.

@schneider_EF @Marc__Watkins @james_s_bedford @thedavenelson Good question. So far, mine have been writing specific and in my classroom. There isn’t much back and forth. What has worked for me so far are attempts for extremely targeted feedback in writing.

@TheBotBrothers @Marc__Watkins @james_s_bedford @thedavenelson Interested whether you came across more deeply rooted "habits" of GPT i.e. to be conciliatory suddenly or evading conflict or falling back into being an AI-assistant when receiving tough questions in a dialogue.

The shocking (but also totally not shocking) point raised at last night’s OpenAI Education Forum were the majority of their global 250 million weekly users are students. Their CFO noted a 90% up tick in ChatGPT usage in the Philippines once the school year began.

@Marc__Watkins @james_s_bedford @thedavenelson I’ve had some success with this via personal GPTs or the Magic School Wrapper of personal GPTs. But it takes some time to write, contextualize, and debug the prompt. A student said I digitized myself. Most teachers do not have the time or knowledge for this.

@james_s_bedford @thedavenelson That’s pretty much what Mollick is attempting to build and open source—customized tutors.

@Brehove I think it is a step in the right direction. Writing teachers are more like math teachers now as showing your work is as important as the answer.

I see that Grammarly Authorship is finally turned on for me in Google docs. Has anyone started using this in some of their courses to help resolve AI policy violations? How wide is the rollout? I've been having challenging discussions with some faculty I supervise around authenticating student work and I wonder if this can help in certain situations. I recall @AnnaRMills having some thoughts on this before?

Super specific bit for AP English Language and Comp 2023 Rhetorical Analysis. It will give focused, specific feedback one point at a time and ask student to revise. chatgpt.com/g/g-x2nHG9zSu-…

@AnnaRMills Anna, Are you using Lex with students or do you know if students are?

OpenAI just released a new interface for working with drafts of a document; you can now chat about the document and edit the document side by side. We've seen elements of this in Google Docs with Gemini and Claude though Lex.page leads. openai.com/index/introduc…

@Marc__Watkins @RosenzweigJane A good point. I was thinking more about how students can now slide a fader up or down change their word choice and syntax. As a writing teacher, I’m scratching my head and thinking how to approach this with students.

@RosenzweigJane Did you play with the reading level? That's one of the more upsetting aspects to me. Developers ship these changes and don't take into account any of the possible long-term consequences to skills, like reading.

@Marc__Watkins It’s a simple matter for anyone to upload an out of copyright text downloaded from Gutenberg and have Notebook generate the podcast, notes, questions, etc. I can also imagine high school/university students uploading class content and using it to “learn.”

NotebookLM's podcast did not materialize out of thin air. Google Labs has been working on the experimental feature for quite some time in a project called Illuminate. It's currently locked to only creating podcasts about research papers form arXiv, but they have used it on lit.

Free read-aloud software recommendations from a listserv I'm on: ElevenLabs Reader app, ReadAloud Chrome extension, and the built-in voice in the Edge browser. Haven't tried them yet...

@markcmarino She is right on. The feedback is limited and average unless you know the specific writing/thinking dilemma that class is having. Then, you can write a bot that will work that.

Check out what my colleague Patti Taylor wrote about the limits of writing feedback by generative AI bots, published in Inside Higher Ed. Such an insightful critique! insidehighered.com/opinion/career…

@AnnaRMills @jtneill @tbertramgallant @emollick @DrSarahEaton Not at all. I just had my students write without AI today and they are working on voice and style using some model essays by Doyle. They understood how AI would interfere with that process.

@jtneill @tbertramgallant @emollick @DrSarahEaton Is a teacher using Google sometimes but asking students not to use it to answer certain homework questions hypocrisy?

If AI misuse needs to be addressed but we don't want to police or use faulty AI detection', what's next? Step 1 is pedagogy that supports intrinsic motivation. But that's not enough. We need to make it less tempting to use AI. Writing process transparency makes sense.

As always, useful for writing teachers and teachers who use writing to assess learning!

We're sharing an update on the advanced Voice Mode we demoed during our Spring Update, which we remain very excited about: We had planned to start rolling this out in alpha to a small group of ChatGPT Plus users in late June, but need one more month to reach our bar to launch. For example, we’re improving the model’s ability to detect and refuse certain content. We’re also working on improving the user experience and preparing our infrastructure to scale to millions while maintaining real-time responses. As part of our iterative deployment strategy, we'll start the alpha with a small group of users to gather feedback and expand based on what we learn. We are planning for all Plus users to have access in the fall. Exact timelines depend on meeting our high safety and reliability bar. We are also working on rolling out the new video and screen sharing capabilities we demoed separately, and will keep you posted on that timeline. ChatGPT’s advanced Voice Mode can understand and respond with emotions and non-verbal cues, moving us closer to real-time, natural conversations with AI. Our mission is to bring these new experiences to you thoughtfully.