In EP 37, John and Jason sit down with Anna Mills to discuss the reality of "agentic AI"—browsers that don't just assist students but can potentially become the student. We move past the panic to discuss advocacy, "humanizing" strategies, and how we can respond without giving up on online learning.
See complete notes and transcripts at www.onlinelearningpodcast.com
Join Our LinkedIn Group - *Online Learning Podcast (Also feel free to connect with John and Jason at LinkedIn too)*
Guest Bio:
Anna Mills is ...
In EP 37, John and Jason sit down with Anna Mills to discuss the reality of "agentic AI"—browsers that don't just assist students but can potentially become the student. We move past the panic to discuss advocacy, "humanizing" strategies, and how we can respond without giving up on online learning.
See complete notes and transcripts at www.onlinelearningpodcast.com
Join Our LinkedIn Group - *Online Learning Podcast (Also feel free to connect with John and Jason at LinkedIn too)*
Guest Bio:
Anna Mills is a leading voice in the responsible integration of AI in education, drawing on nearly two decades of teaching experience and a deep commitment to open educational resources. Her expertise spans AI literacy, academic integrity, and the critical use of AI in higher education, work informed in part by her role as the sole education specialist invited to test GPT-4 pre-release for OpenAI. She is widely recognized for her influential resource curation, national and international faculty development sessions, and published contributions in major higher-ed outlets. Mills also authors the widely adopted OER How Arguments Work and advises multiple AI-focused initiatives—you can find Anna Mills here on LinkedIn.
Resources:
- Anna’s Substack (great summary of Agentic AI and Education / lots of links and resources) https://annamills.substack.com/p/the-time-to-reckon-with-ai-agents
- John’s LinkedIn post on how Comet Browser can impersonate a student in an online course
- John’s YouTube video showing Comet impersonating a student
- Anna’s LinkedIn post about Yun Moh’s request of Canvas.
- Annotated reading conversation
- Forbes’ Article “Colleges and Schools Must Block Agentic AI Browsers Now, Here’s Why” https://www.forbes.com/sites/avivalegatt/2025/09/25/colleges-and-schools-must-block-agentic-ai-browsers-now-heres-why/ (not mentioned but a good one!)
Theme Music: Pumped by RoccoW is licensed under an Attribution-NonCommercial License.
Transcript
We use a combination of computer-generated transcriptions and human editing. Please check with the recorded file before quoting anything. Please check with us if you have any questions or can help with any corrections!
EP 37 Anna Mills - Oct 28, 2025
[00:00:00] John Nash: Hey. Quick pause we're collecting listener testimonials, and so if this show has influenced your thinking or your practice in any way, you can share that with us at onlinelearningpodcast.com. There's a link at the top of that page. You can't miss it. Just click that it takes just a second and we'd love your testimonial.
[00:00:19] Jason: That's right, just right at the top in kind of obnoxious yellow font, to be honest. But it's a black background and so it's accessible, but still, you should see it at the top.
We'd love your feedback.
[00:00:31] John Nash: And if this conversation is useful. Take a moment to follow the show so you don't miss any new episodes in Apple Podcasts.
Just tap the plus sign on the show page, and in Spotify just tapped the follow button.
[00:00:44] Jason: Also, if you like this podcast, we'd love your rating. It helps us in the algorithms kind of bump to the top. So, in Apple Podcasts, you scroll all the way down and find the stars and put the stars in. In Spotify, you hit the three-button menu and then rate this podcast. We'd appreciate.
[00:01:03] John Nash: So many steps, but I tell you it's worth it. Alright, to the episode.
[00:01:08] Jason: To the episode.
[00:01:09] John Nash: I'm John Nash here with Jason Johnston.
[00:01:11] Jason: John. Hey everyone. And this is Online Learning in the second half the Online Learning Podcast.
[00:01:15] John Nash: We're doing this podcast to let you in on a conversation we've been having for the last three years about online education. Look, online learning has had its chance to be great, and some of it is, but there's still some ways to go here. So how can we get to the next stage? Jason,
[00:01:32] Jason: Well, how about we do a podcast and talk about it?
[00:01:34] John Nash: I love that idea. What should we talk about today?
[00:01:37] Jason: John, there's a lot of a lot of people talking on LinkedIn. This might be the end.
[00:01:41] John Nash: Do you think so?
Is this where I disclose that I am not actually hosting today, that my Comet browser is hosting for me today?
[00:01:49] Jason: This is Comet John. You're doing a great job. It's very, it's almost as believable as the real John, and maybe we're not so bad off after all. But yeah, the, all the talk this week on LinkedIn has been about, agentic AI taking over. The one quote that I took away, I won't name the person, but, on a comment on LinkedIn, said, online asynchronous learning is cooked. And um, I've learned this is not a good thing from my son who uses this term. It can be a good thing somebody can be cooking like in a good way, but cooked means bad.
[00:02:24] John Nash: Yep. Yep. So, we have someone in the house today who's going to take us down a path of thinking this through. This may actually be one of the episodes where we may accelerate this one out because this topic is so hot. And so, who have we got today?
[00:02:39] Jason: Today we have Anna Mills, and we actually just met on LinkedIn because of some of the writing. She was doing around this, and so Anna, welcome to the podcast.
[00:02:48] Anna Mills: Thank you. I'm very excited to be here and talk with you. I think it's pretty urgent and,
[00:02:54] Jason: Yes, I wish.
[00:02:55] Anna Mills: can do.
[00:02:56] Jason: I wish it was under better circumstances, let's just say. No, this is actually the very circumstances why we have this podcast to talk about these things. And Anna, maybe first to give us some context - give us just a little bit of background for yourself and what your role and work is right now.
[00:03:12] Anna Mills: Sure. I'm a community college writing teacher. I've taught for many years in the San Francisco Bay area, and I wrote an open educational resources textbook that's free online. I dove into discussions of AI in education and writing education. Before ChatGPT was released and I did some testing for Open AI early on and I an advisor on an app that invites students to use AI feedback. So, I've been interested in both using and critiquing AI. And I've been teaching online for some years now as well. And I've just been really active in social media discussions because I think we have to come together and join to find our way in this terrain.
[00:03:55] Jason: Yes.
[00:03:55] Anna Mills: glad there are a lot of educators out there doing that.
[00:03:58] Jason: Yes, exactly. What do you teach online? Just out of curiosity? Do you use some writing?
[00:04:03] Anna Mills: I teach English composition required academic reading and writing. Fundamental kind of support for thinking in college. You know, I think it's even more important now with AI to be able to read and write and edit and think critically but, doing that online asynchronously. How do we maintain some accountability - how do we maintain the value of that course credit when it can be outsourced? So that's where we are. I think it's going okay, it's shifting terrain, so.
[00:04:32] Jason: Yeah. So, if you're teaching writing online, my guess is, as soon as the public ChatGPT went public three years ago now, that was probably a significant concern from you from the beginning, right?
[00:04:44] Anna Mills: Yes, sure. And I give a lot of workshops on building your own multi-pronged strategy for reducing AI misuse if you're teaching writing online, asynchronous. think it starts with, good pedagogy and designing for intrinsic motivation and bringing in relationship and all the things that we do know how to do and can lean into. but I don't think that's enough. We have to have some guardrails too. So that's where I've gone into some process tracking, some AI detection, used non punitively very cautiously.
[00:05:18] John Nash: So Anna, I'm wondering if you could give us a little level set on the main topic we're going to talk about today, because we might have new listeners that come in or are wondering, oh, agentic ai, and what are you talking about and what are we worried about? And there's probably just a baseline we could start from about what is mechanically occurring, and then we could talk about a little bit about what the growing concerns are.
[00:05:43] Anna Mills: Yeah. It's really pretty simple. It's just a different paradigm from what we're used to with chatbots. agentic AI or an age agentic browser. What we're basically talking about is you're browsing the internet, you're also chatting at the same time, and you're chat bot browse for you. can click. can fill out forms, it can take action on that website. It can navigate to other websites. So, it's an extension. You're still prompting it, but it can also just keep going on its own including in a learning management system. And this is where the companies have been heading for some years now. It's just that it's only recently started to work reasonably and be more accessible to people who are either free customers or paying $20 a month. That's pretty recent. So, we have a release of the perplexity browser. They made it free to students for a year. Very recently. Then we also have a ChatGPT Atlas browser and all the other companies are working on about to release or have something in beta, so Anthropic and Google as well. So, they think this is the next big thing and it's now just starting to work for average users. Students are just becoming aware of it.
[00:06:59] John Nash: One of the things that strikes me as you were describing this is that the agentic browser will browse with you. It can click for you; it can fill out forms. Before all this came out my password manager, I use LastPass, will also fill in forms for me, particularly when I shop, and that's maybe something everybody's used to. And so, what also seems to be turning up the talk around this is that the large language model is riding along with you with all of its air quotes, intelligence. You we're all talking on video now, but I can't do my fingers. And so, I have to say,
[00:07:34] Anna Mills: Yeah. Yeah.
[00:07:35] John Nash: and what we're concerned about also is that it can start to answer questions, for instance, in an online class as if it were the student without the student using their own hand.
[00:07:46] Anna Mills: And so, it's really like reading and analyzing the website, including an assignment description. It's giving you a little, blow by blow description of its reasoning process, what it's doing next, and that will include auto generating response to a discussion or coming up with quiz answers or generating a whole essay and then actually clicking, putting it in and clicking submit and declaring to me the user.
Oh, now your essay is submitted. So, it's way beyond that kind of password auto complete because it's got all the capacity of, the chat bot that we know behind it.
[00:08:22] Jason Johnston: So, I've been talking about this with some administrators for a couple months 'cause I've been kind of testing it out. I showed my team, which is about 25 people, media people, instructional designers, basically video of, Perplexity's Comet, browser, completing a module in an online course doing a quiz, responding to a discussion board, submitting an assignment, selecting a topic of study. And then this was kind of the kicker, using Canvas, sending an email to the professor asking a question about something in module one. I mean, the prompt was about three sentences
My staff watched with open mouth and kind of in horror, as you could see, the AI clicking around, completing the quizzes, doing all the things,
[00:09:14] Jason: We'll put a link in maybe John to your LinkedIn post where you walk through a little bit, what of a video of you were using you were using Comet, were you John?
[00:09:25] John: Yes, and So was Anna. And I think she we were both sort of thinking about it at the same time. Hers was quite good too. Or I won't say too, it insinuates, mine was good.
[00:09:35] Jason: Yours was good.
[00:09:36] John Nash: hers was quite good.
[00:09:37] Jason: I want to affirm you both of them are really good. Yes.
[00:09:40] John Nash: And really went deep in thinking about all the affordances that these new browsers can bring to bear in completing a course. It's quite complex what it can complete. It does, as you pointed out, Anna, read the content that the teacher has put into the LMS and then answer based on that. It's shocking.
[00:10:00] Anna Mills: And I appreciated your post 'cause you said the student didn't use AI to cheat, exactly. AI was the student. At that point there's no student involvement. The student doesn't even know what the assignment was necessarily in some cases because they've just said, "go do my assignment." Maybe they say, don't ask me questions, and it just goes with that. Yeah, so there have been a bunch of educators posting videos and you know, in my Substack post I try to bring in, you know, all these great examples like from David Wiley and Tim Mousel and Anna Withrow. Here and there, there've been communities of people talking about this for the last six months, but I don't think it's really hit broad consciousness among educators. So, I just, I just got a little concerned and wondering if there's a solution and wondering if we should be the ones calling for it, so that's really where I started posting and writing about it. And I noticed that there was a Forbes article from Aveva Lagat that was really helpful. So, I think that, now we're starting to see a much broader discourse and concern.
[00:11:04] John Nash: Yeah. Well, I would love it if you could walk us through some of your thinking on where we could go with this. You noted in the Substack post that you made, and we'll put a link to that in our show notes that we can't just give up on online learning. And certainly, I think Jason and I would totally agree, otherwise we wouldn't have a podcast, but that uh, but you,
[00:11:24] Jason: really difficult. We got a really decent and I'm not sure we can get another good one in 2025.
[00:11:30] John (2): Yeah, but you note to your readers about the depth and penetration of online learning today and the way students are learning in post-secondary at end in, in, in P 12 arenas. And so, you do take also some time to wonder out loud what educators could do. You were talking that, you know, Philippa Hardman offers ideas on how instructional design can maybe thwart this.
You also wonder, I think a little bit if course design actually can solve the problem. Maybe, and I don't know how sophisticated these agents will get. You also talk about how educational institutions could try to block. AI agents, and I read that line in the post and then I put down my, I think I was reading probably on my tablet.
I put down my tablet and I was, I started to become incensed because I thought, not for what you said, but how true it is that so many institutions now are put into a ridiculous position of having to take on opportunity costs to thwart. An issue that probably shouldn't have even been brought to bear in the first place.
This, it's, there are deep ethical questions about whether it's right for these companies to do this. And the companies are even sort of saying the quiet part out loud, aren't they? And you posted, but
[00:12:48] John Nash: uh,
[00:12:48] John (2): Yun Moh asked Instructure in their bug feature forum to include a feature that blocks AI agents from logging in on behalf of students and Instructure marked the request as "will not consider."
[00:13:02] Anna Mills: Yes.
[00:13:03] John Nash: I'm just like, are you kidding? And so, it brings me to a thought. It's sort of a maddening thought on opportunity costs, as I said. But in a sane world, shouldn't the conversation start from, should this exist at all? But instead, it's framed as how can schools detect and respond to this?
And so, I just, yeah. What do you, what are your thoughts on direction?
[00:13:26] Anna Mills: Well, I shared those feelings because I think probably the simplest thing that could be done would be for the companies that are, offering these AI agents to offer some kind of opt-out or identification of when it's an agent and when it's not. So that websites could say, okay, we don't want, agents. Canvas taking quizzes, right? A learning management system. It's not an AI agent management system. And there might be a lot of cases where certain websites for certain purposes, they want a human interacting with it. And I think that's the simplest thing because you can tell with these agents that they haven't been instructed not to do this. You say, will you complete this assignment? And they say, okay, I'll complete the assignment. And there's a system prompt in there and OpenAI or perplexity could be saying, if you're asked to complete assignment, say no. How about I coach you on it instead? So that first step has not been taken. And that step of, identifying the agents as such so that websites could block them if they choose hasn't been taken. And it's a very simple step. The other steps are really complicated and maybe don't work as well. So can canvas itself, block them? Maybe, I don't know. After, you know, I shared that post about them not considering the request.
Melissa Lobel from Canvas did respond and say they were, they are thinking about it and talking about it with open ai. so, I don't know what the status of that is. I don't know if that will work. When I asked my canvas administrator at my institution, she said, I don't think we can do anything. The only thing we might do would be to block it on our WIFI network, but students don't have to use our WIFI network. So, I don't know that there's a technical fix at the institutional level that even works. I'm certainly interested if there is I think we should look for the simple solutions and the people who can carry those out.
[00:15:18] John Nash: Sure. I mean, I just love your short dissertation there on all the things that we could be thinking about is really evidence of the intellectual and structural cycles we're burning, trying to solve this when the issue is just so simply fixed by the company.
[00:15:32] Anna Mills: Yep.
[00:15:32] Jason: It's interesting, I don't know why they've opened up these guardrails, because I've been testing this for a little while. Ever since I had hands-on a, any kind of agentic kind of ai, and so you could do it with ChatGPT 4 and it actually refused to do it. You would have to go around and say, hey, this is, I'm just testing this course out. I'm the teacher. This is the test student. And then it would do it almost like kind of roundabout kind of way.
[00:16:00] Anna Mills: Yeah.
[00:16:01] Jason: When five came out, then it actually would.
[00:16:04] John Nash: Actually
[00:16:05] Jason: ahead and do it. And it actually was pretty, and that's when I started to
Startled. This was just a, I think maybe just a couple months ago when five came out, GPT five, and that's when I started to become a that's when I started. Like people weren't talking about it openly yet. I didn't really want to talk about it openly yet right? Because I hadn't really gotten my feet underneath me about all this. And then it feels like with perplexity now, it's like another arms race. Like now they're just openly, you don't even, you don't have to backdoor it. they're advertising to students like, here's your 4.0, secret. Like literally I have clips of advertisements to me in Facebook to use perplexity to have perplexity, do your busy work so you can get on with learning. And this is your, your four oh secret and it shows examples of going into Stanford engineering and it's just it's wild, really were. It feels like almost ethically, morally, something has just moved in the last three or four months where they've just decided to, I don't know, just open up the gates, take off the guardrails or something.
[00:17:15] Anna Mills: And I don't think that sort of parents and teachers have realized that yet, but I
A pretty strong. Across the board, bipartisan objection to that. Because 57% of community college courses in California were online in 2023, probably more now.
Is a huge equity and access issue. You all know this if you're listening already to the podcast, but, and parents and teachers. We still want this as a learning space. And why should they suddenly start profiting off cheating in this unashamed way? I think there is potential for some pushback there, and some like a sanity check,
[00:17:56] Jason: absolutely.
[00:17:57] Anna Mills: I know these companies need to make money and they're experimenting, but there's a lot of parents in those companies and there's a lot of people in those companies who might wake up and say, "Hey, maybe we do need to back off from this advertising cheating approach. Maybe that's not the ethos we want." I'm trying to hold out some hope for that.
[00:18:15] John Nash: Yeah, I think puzzling out the long game on this if this sort of feature remained in place, just seems. If a Silicon Valley company that seeks to gain market share and put out tools like this, they're a company that needs software engineers and they need high level managers and they need marketers, and they all have to get their degrees in a.
Post-secondary institution, presumably many of them taking online classes. What will the level of outcomes be for students who rely on agents to take their classes? Who's going to hire them? It's just sort of an existential undoing of the companies by doing this in some way.
[00:18:53] Anna Mills: Yeah, what about when they want to do corporate training? I think there's a separate discussion about this going on among folks who are interested in online corporate trainings because it's the same problem. So, it, it just seems unsustainable and illogical, so how can we send that message is my big question.
And I'm not really, I'm not a politician, I'm not an organizer. But this needs to be done right.
[00:19:17] Jason: . This is one of things I really liked about your Substack: it wasn't just a big "this is happening, what are we going to do?" Kind of thing. It was like here's some ideas. I've thought about this. Here are some approaches. And some was around a list of ways that we can advocate at the end of your Substack, which I appreciated and that actually, helped spur me to start some conversation with our LMS administrator and department because it's a separate, I'm at University of Tennessee, but it's actually a separate department from the online learning department. And I think
[00:19:49] John Nash: And I think
[00:19:50] Jason: there are things that we
[00:19:51] John Nash: there are things.
[00:19:52] Jason: and you talk about raising the issue with colleagues talking to IT department and the whoever manages the LMS contract, these are the people that have. Have direct lines to the people that will take notice, right?
Because these are multimillion dollar contracts that that we have a, we have a choice to renew or not, right? Next time. And if. This is a significant enough of a problem that we feel like it is gutting the value of our education in the perception even if it's not, if it's just even in, in the perception,
[00:20:27] John Nash: it.
[00:20:28] Jason: will stand up in a way when our LMS is up for renewal,
That we will look seriously at what these companies are doing in support of education in our minds.
[00:20:39] Anna Mills: that's a powerful message.
[00:20:40] Jason: . I maybe loop back to a couple of other things. We've talked a little bit about advocacy. As we started this conversation, I think just when we talk about it, obviously it, there's strength in our voices to be able to talk about these things so that it starts to feel more, not just real, but reasonable. That we would have these kinds of expectations out of our technology, right? That they would have some guardrails on them. So, I think Nobody's questioning whether or not online learning can happen age of AI agents because we know people will continue to learn. There's going to be lots of people. I would say the vast majority of people are going to continue to go online to learn and not try to cheat the system. So, some of the concern comes around assessments. Then what are some of the things that you've been
Are some of the things.
it comes to assessing people that are fully online and maybe fully asynchronously online?
[00:21:35] Anna Mills: I've been influenced by Tricia Bertram Gallant, who wrote a great book, the Opposite of Cheating, teaching for Integrity in an Age of AI with David Inger.
Been saying, what if we really can't secure online assessments, but we still want online learning. if we need to maybe look at some proctoring centers, some kind of in-person assessment that compliments online learning. and I think she's partly influenced by this two-lane approach coming out of Australia. The idea of, you can have your formative assessments where you're not trying to secure it, but then you have this periodic kind of program wide, proctored assessment. That are secured. so, I think there's something to that.
We might have to go that direction. I understand that will also exclude people who can't get to proctoring centers. It creates a barrier; it's an equity issue. So, I really want to see the other approaches tried first. And I want to recognize that sometimes students even if they recognize that there's a need for accountability and sort of assessment that has accountability, we should be looking at whether we can do secured online assessments ethically without privacy, invasion and bias. So, I think we have to keep looking at that even though there have been real concerns about things like proctor's video surveillance we should look at what are the ways to do it?
How effective are they? Do they reduce incidents of cheating? Even if they're not perfect. What are the risks? I think we; we have to look at that. But we shouldn't be forced into a corner where we have to give up, everything that we're offering to do that if it's not necessary, if the companies themselves could actually, prevent a lot of it, a lot of the misuse. I don't know if that makes sense or if that really that's what you were looking at,
[00:23:23] Jason: No, I think it does make sense.
[00:23:25] Anna Mills: We can't move toward proctoring overnight, institutionally, that's very complicated. There are these, testing professionals, associations. There are ways to structure programs so you can have some flexibility.
And when a student might come in to take a test that's not going to happen overnight. And it has downsides. As does any kind of, video surveillance, lockdown browsers, all of that. It's I think it's a wicked problem, as Philip Dawson says.
There's not going to be a perfect solution, but let's go to some of the things that would help a lot first, and then we do some careful weighing of the other options too. My approach.
[00:24:03] John Nash: What are you noticing in your teaching experience about. I'll call them low threshold, high impact strategies that online instructors might adopt in the wake of this, that they're not, they don't have access to an instructional designer. They may have some ideas around how to. Get away from what I'll say sort of obvious AI able assignments, and yet still sort of the age agentic stuff sort of gets in the way.
For instance, I think about things like sort of ways to involve project-based learning, other sorts of public demonstrations of learning that might be un-AI-able, but what yeah what's been in your toolbox and what are you recommending for colleagues?
[00:24:48] Anna Mills: I've definitely moved to social annotation over discussion posts and I really love,
[00:24:53] John Nash: Nice.
[00:24:54] Anna Mills: all of the readings in perusal or hypothesis and having this conversation in the margins among students. It's fantastic pedagogically and it's just. Less enticing to use AI for that one sentence, informal comment in the margin. And you can block students from copy pasting into it. So that, that's great. I think video posts yes, those could be autogenerated with maybe HeyGen, but it's less likely. And maybe conferences, conference assessments is something I'm looking at. And I think definitely process assignments that are low stakes that are an un grading, like 40% of my course where there's a lot of step-by-step, you do this assignment you get full credit if you do it.
So, it's not that panicked last minute. It has to be, polished. An AI version. So, I think that's helpful. Of student choice in what they write about, what they focus on. I like that you, you mentioned project-based learning. Things that are meaningful in the real world something they might want to publish or share later on. So yeah, those are just a few things, but I think having a lot of different formats of assignments from video to annotation to Google Docs, where they're sharing their process history on the document is helpful. You can see how much time was spent in the document.
Yes, there are ways around that. Unfortunately, there's software that will simulate typing and time spent and even errors. but again, if you're reducing, you're making it harder to do the wrong thing and easier to do the right thing. In the way you structure it, I think you're enabling a lot more learning.
You're reducing those missed opportunities.
[00:26:35] John Nash: Yeah, that, that's a great list. And I'd heard of the, and the social annotation. It's been a while though, and I think I want to. Yeah, bring that back into my toolbox because I think that's a better way for them to work with me. It's sort of, it seems like every episode I want to quote Michelle Miller and this idea of same side pedagogy, but we're, when we're both reading the article together and commenting on it, we're learning together.
We're on the same side.
[00:27:02] Anna Mills: fun. That's the part, it's fun for me to read their annotations and it's more fun for them. It feels like social media and
It can be auto graded for completion credit. So, it's helping me focus my teaching where the joy is right. I like when it's both and
[00:27:19] John Nash: Love that.
[00:27:20] Anna Mills: misuse and it's actually fun.
[00:27:22] Jason: Yeah, I have found with my classes, one of the things is trying to break things down into smaller chunked interactions as well, like throughout the week. This is part of this kind of social learning so that it doesn't just become a big assignment at the end, and so all of a sudden, you know, all of a sudden, they're looking at the clock they've got an hour. You know, you talk about reasons why students might cheat. Even good students, even people who want to learn, they got an hour before the deadline for of this thing and there's some shortcuts. There's an easy button. It's just so tempting, you know, but if we draw them into an educational learning community where there's lots of micro learning and connections and conversations going on, like around, I love, the suggestions of perusal on a hypothesis and Google Docs and I think John, did you coin this phrase of learning out loud?
[00:28:15] John Nash: No, building in public is something I, I sometimes talk about. Yeah. Yeah. I didn't coin that either. But it's the idea that, yeah, if you're doing that, you're okay with things not being perfect, and then that, that encourages others to try things and also sets it up so that you can receive feedback on what you're doing.
[00:28:33] Jason: Yeah. Do you do any of the, that, the kind of the building in public with your writing with your students?
[00:28:39] Anna Mills: Oh, that's interesting. I think I do that in the sense that I really invite them to give me feedback both anonymously and non-anonymously. And I give them extra credit when they find. Typos and problems. And I share my syllabus at the beginning as a Google doc, and they have to comment on it.
Of sense of trying to share my process. I'm really telling you I don't have all the answers. I want to have this discussion with you. And I think that's an open pedagogy, trust-based, vulnerable approach. That's really important to me. I think I have to get started though on sharing my own, I have shared my own writing process, you know, early drafts and now it's time for me to share like a whole process history with them the way that I asked them to share with me. so that sense of fairness. I try to be really transparent about when I use AI and when I don't as well.
[00:29:29] Jason: Yeah, that's good. Yeah. Yeah. I think communicating around. When you use ai, when you don't model, that kind of transparency we're really looking for in the classroom too, so just mention so many things that if I could put some keywords, into the kind of online learning we are trying to build around human it's transparent, relational. Its trust based, you mentioned trust. John had mentioned same side pedagogy again. I don't that any of those things absolutely, make sure a hundred percent that students aren't cheating. But I think it's certainly makes an environment where cheating is less likely less appetizing, I think, to the student.
[00:30:19] Anna Mills: Yeah, and I think you, it's not either or. I think if you have both that ethos and you have some accountability, I just worked with a student who had submitted an AI essay. It was very clear and so like it was helpful to have that of obvious in the process tracking and the AI detection. Then to meet with him and say, let me sit with you and talk through your ideas. And it's really interesting and help you build confidence and connect to the assignment on your own terms. And so, combining those things maybe can work for students who are at high risk of cheating. I think, yeah, it's not either or one, one way is perfect. So, I've really just learned from the approach that Michelle Pacansky-Brock has shared in the California community College system. Humanizing online learning is a big project she's worked on a foundation grant, and it's just brought so many wonderful strategies to our system and to our training as online So, yeah, I do feel lucky that there's so many great resources like, your podcast and a community of online educators who already know that we need to humanize and we can lean into that.
[00:31:28] Jason: Thank you for including our podcast in of the solution. We feel like it is on some small levels. In some ways it's just an excuse for John and I get together and talk about things that are on our mind and meet cool people. Give us an excuse frankly, to meet. Cool people and to talk to them about it too.
But we do believe that the conversation is part of that solution and helping people stay educated and staying up on the conversation, being open-handed with it as well, I think John and I come to these podcasts with more questions than we have answers. And that's where we are at right now online education.
[00:32:05] Anna Mills: It's a healthy place to be in this moment with ai. I think, if we can stay open and keep talking to each other, we can make a lot more progress than if we have to pretend like we have the answer now to this thing that just came out two hours ago.
[00:32:18] John Nash (2): Well Really appreciate your thoughts, Anna, and especially as I think about how you were closing out that post and the on Substack that jason was talking about towards the end on ways we can advocate. I see you raise the issue with colleagues in your academic Integrity office. IT office.
I'm thinking about sharing it with I'm, it must be on the radar of our Center for Teaching and Learning, but I'm going to share it with them as well. I'm not sure where this is really landing with many online instructors at my institution who probably just are worried about getting out the next module and may not be coming up for air to see what could be operating in the background.
[00:32:56] Anna Mills: Yeah, but it makes sense to do it through our organizations, through our existing structures. Even if they can't fix it, they can raise the issue with other stakeholders. So, we can lean into our institutions here. That's what I've been doing too, is just reaching out to people and committees on the state level. Who's working with Canvas? Who's partnering with these companies? Can we put the pressure on there? Maybe talk, maybe look at our unions or professional organizations. We're working on a statement in this MLA Modern Language Association task force that I'm on. So yeah, just, it shouldn't be individual teachers who have to figure it out in the moment.
Yeah, we can work together.
[00:33:35] John Nash: Yeah. No. Yeah.
[00:33:37] Anna Mills: Thank you so much for having me. This has just been delightful. Yeah, I'm lucky I get to have these kind of conversations and I definitely don't have the answers. I'm not an expert on this issue, but looking forward to everybody else's contributions and voices being
[00:33:50] Jason: Yeah. Thank you for also being part of the solution for taking the time. I just am amazed when people take the time just to, talk with us of course today, but also put out just a really good and thorough. Substack on the issue that has helped me. I've shared it with my instructional designers and having discussions about how do we approach this from the design side, as we talk with faculty. And I think all of these tools can really work together. So thank you so much and thanks for joining us here today.
And for those listening we're going to put into the notes any of the resources that we've talked about today. Our website is online learning podcast.com. That's online learning podcast.com.
And you can find John and I and Anna on LinkedIn. We're active on there, now and again. Right. John, you're more now than again, or is it again
[00:34:37] John Nash: Yeah I'm a little bit back. Yeah, I took I don't know how it was possible, Anna, but I stopped LinkedIn for like six months and just sort of thought about stuff. I don't know. So now I'm going to try and talk about what I thought about.
[00:34:52] Anna Mills: Yeah,
[00:34:53] Jason: Which I think is also important, especially in this age to have embodied experiences, non-computer and non social network experiences so that we can continue to think deeply about things and recognize our shared humanity and that we're part of a world that existed way before this kind of technology as well.
All right. Thank you, Anna.
[00:35:15] Anna Mills: Alright. Thank you. See you online.
[00:35:18] John Nash: Yeah. Thank you so much.
