Estimated reading time: 8 minutes, 4 seconds
To unpack this, we should first maybe unpack crticial, AI, and literacy.
By critical, I mean this in multiple senses of the word. One is critical as in critical thinking, as in skepticism and questioning. The other is critical as in critical pedagogy, so focusing on social justice dimensions and inequalities something may exacerbate, reproduce or create. Another is critical, as in, critique it for its potential harms, and critique the credibility/accuracy of its outputs/outcomes (this is I guess can also come from a combo of the other two criticals anyway). So kind of like what information literacy, media literacy and digital literacy do – I think it’s closest to digital literacy in the sense that info and media literacy have not historically maybe dealt with algorithmic literacy, but I’m thinking it’s going to be part of all three of them anyway in future and there is loads of overlap anyway!
AI as in Artificial Intelligence more broadly, not just chatGPT, with all we have learned from how it works, and how it has impacted us historically – whether machine learning type AI or some other type. Recently, a faculty member at my institution reminded me that maybe we need to remind people that it’s not truly “intelligent” in any sense of the word, because we know it’s just a statistical/probabilistic model. True. But the term used around it is AI, so it would be good to question how this term is misleading, but odd to stop using it, if that makes sense? I guess that is also part of AI literacy – bring it down to size! Recognize how its name itself is mesmerizing and potentially biasing us to trust it more than we should?
Literacy as in beyond the basic skill of how to use something, but beyond that into the capacity to know when, where and why to use it for a purpose, and, importantly, when NOT to use it. That’s the basics.
Now how would we focus on teaching critcal AI literacies in this fast-moving landscape we find ourselves in? It’s definitely dynamic and something I don’t think many people hold authority on at the moment. To understand both the technical dimensions fully and the social and ethical dimensions. No one expects every professor of math, history, economics, and literature to understand all of this fully, but many will be impacted in their teaching and in their profession. So, while not many educators have reached expertise in this, it is something we can all learn with our students and also with other educators within our institutions and around the world.
What I’m sharing below is not like my neatly packaged response, but bits and pieces here and there from things I’ve written about, workshopped and presented about, and taught, related to Critical AI Literacy.
What to read on the journey to critical AI literacy?
General AI Ethics Considerations
When Anna Mills and I gave an Equity Unbound session on critically incorporating AI into teaching and developing AI literacy, we recommended the following readings of varying lengths:
- If you have 2 minutes: Check out this infographic on AI ethics a post by Leon Furze
- If you have time for a deep dive: Read Autumm Caines’ article on conversations we should all have with our students before (if) we use AI with them (including exploitation of Kenyan workers in the process) (she focuses on labor and data privacy) or read more from Leon Furze on bias, environmental impact, and truth
But of course critical AI literacy predates ChatGPT and all of that, and I’ve been teaching some kind of “algorithmic” literacy to my students in my digital literacies class for a long time. We don’t read books for this because it’s not the main topic of my class, but if we were reading books, I would have recommended (keep in mind my course teaches other things so algorithmic literacy is part of but not central to it).
Readings/videos on inequality/oppression created, exacerbated, or reproduced by AI/algorithms:
- Safiyya Umoja Noble’s Algorithms of Oppression. Since we don’t have time to read it, students either watch this TED Talk by Safiya Noble OR read Engine Failure interview with Noble and Roberts
- Zeynep Tufekci’s Twitter and Tear Gas – but in my class we watch this TED Talk by Zeynep
- Ruha Benjamin’s Race After Technology. We see this video promo for it
- Cathy O’Neill’s Weapons of Math Desctruction.
- Joy Buolamweni’s TED Talk “How I’m Fighting Bias in Algorithms” or more recently, the Coded Bias film on Netflix.
- Shea Swauger’s Our Bodies Encoded which critiques algorithmic test proctoring.
You get where this is going, right? There are of course many people writing really great stuff around AI these days, and one of my favorites is Brenna Clark Gray – check out her posts in the TRU Digital Detox site which is focused on AI this term. Another person who has been writing a lot around AI for a while is Chris Gilliard and he’s written a couple of things recently, such as this one with Pete Rorabaugh, but do check out other articles, videos and podcasts with him.
How to Use AI in Teaching to Promote Students’ AI Literacy
When I say this, I mean helping students recognizing the limitations and potential of AI – what they gain or potentially lose when using it. So I’m recommending things for teachers to check out, which they can reuse or adapt for their own use, so that even if they don’t have time to look up ways to teach about AI, these links can help.
When Anna Mills and I gave our session, we told participants: If you have 4 minutes: Watch this video about how a teacher incorporated AI into her high school English teaching & how her students responded in this follow up video
I would also check out what Ethan Mollick has been writing on this. Check out the entire substack
I would also of course check out Anna Mills’ AI Text Generators and Teaching Writing: Starting Points of Inquiry
There are also some spaces where someone can learn something like prompt engineering – check out LinkedIn’s Prompt Engineering course on this and this free course ChatGPT for educators by Bron Eager (also her book AI Prompt Phrasebook – available on Kindle Unlimited; got it, but have not read it yet)
Other things I recommend when I give an invited talk or keynote or webinar are these two resources:
Creatively explore speculative futures of AI. Invite your students to read this article by many authors from around the world, reflect, then write their own fictional future story of AI (about their field? about something else) as a way to empower learners as citizens to imagine the future of AI they either strive towards or are afraid of – so that hopefully one day if they’re in a position to make decisions, they’ve thought through this.
I would also recommend engaging with Sarah Elaine Eaton’s suggestion that are approaching a postplagiarism era (and here is my response to her – I don’t agree with everything!)
Check out this crowdsourced 101 Creative Uses of AI in Education (link)
Curate and learn from what other educators are doing around the world in terms of incorporating AI or giving guidelines or boundaries of its use in their classes. We curated what faculty at my institution are doing, for example.
How Does AI Work?
Obviously AI is a wide variety of things, and they work in different ways. I usually focus on “machine learning” (which again, more than one way) because it’s the most, I think, different from how we expect computers to work, because they’re not “programmed” to behave a particular way, but the software “learns” from loads of data.
People seem to find the article On the Dangers of Stochastic Parrots: Can Language Models Be Too Big? useful, but I haven’t read it in depth yet (not sure why!) and someone just sent me this one What is ChatGPT Doing and Why Does it Work?
I’ve never struggled too much with this question because my undergraduate thesis used machine learning/neural networks, so I understand the basics. But for my students, other than orally explaining, I usually let them play QuickDraw then we discuss (I recently added these other two games):
- QuickDraw: QuickDraw
- Real or Fake (text): https://roft.io
- Real or Fake (photos): Which Face is Real
Taking the Socially Constructivist Route?
In all of this, though, I would suggest we all take the socially constructivist route. I don’t think I’m an expert on AI or AI literacy or anything, but I am someone who suggests we work with it this way. Wherever/however it makes sense in your course, at the level of the students, and with whatever energy you’re able to give it, bring in these conversations and listen to what students are thinking and saying. For some people, this may mean showcasing AI to students, for some, just talking about it, and for others, allowing students to experiment and then reflect on how they’ve used it and how it’s helped them. There isn’t one right way, there is the way that works for you within the parameters of your situation.
What does Critical AI Literacy mean to you? Is there something you consider to be REALLY important that I’ve missed out on here?
Featured Image by 0fjd125gk87 from Pixabay (the man holding his face in his hands as a metaphor for looking oneself in the mirror, critical reflection; the wires around his arm and wrist are the technology part, I guess)
As usual you provide a thorough breakdown of terms at play with many useful resources.
I’ve been pondering what is meant when we say something is produced by a generative process. What we do as humans is we read see many kinds of information (training) that we draw upon when we sit down to write or create art. Yet we are not algorithmic? Or is it that we have a purpose or intent?
At the same time the process of how AI content to me, with some technical understanding and attempted readings of academic papers, cannot find an intuitive concept of what a text or image generators actually does when it spits something out.
The new video for Everything is a Remix presents to me some clear and probing questions on what is meant by creating art.
Thus I appreciate your approach of asking students to explore through QuickDraw and the others. I had not tried the where you have to guess if text was human or machine written, I did poorly. There is one like it for images (try search on This Image Does Not Exist) where you try to determine if an image was done by Ai or human.
I’ve wondered too of a flip on the so typical thing done where someone published a post or article only to reveal at the end it was by AI. Ive sometimes tried respond to posts or blogs trying to write as if I was ChatGPT (use lots of sentences starting with “It is important to”) etc, might be an exercise to identify AI styles (likely harder in newest wave of tools).
Well just a long comment meant to be mostly appreciation for what you do so well. And noting too in your featured image credit including too the reason you chose it- some people notice the small things!)
I also noticed your held back on putting links so that your post would moderate immediately and not need me to naturally approve it! Yay! Will follow up on some of what you said here. Find the links I mean haha
This is so helpful! I love how you start by explaining what you mean by “critical,” “AI,” and “literacy,” which introduces key concepts.
I also especially love the way you are curating what faculty are doing at your institution–that seems like a wonderful, simple way to encourage faculty to engage with this.
It would be cool to see a “real or auto-generated” text activity online updated with GPT-4 outputs.
One of the most meaningful pieces. Thanks for taking the time to simplify the concepts.
Thank you so much for sharing this.
This intersection between technology and literacy is something you and others (and myself) have been doing for some time, so I think we have an ability to at least frame the inquiry. Your post here is helpful, moving our thinking along the lines of critical literacy analysis of the AI emergence via Chats (not just GPT, of course)
Kevin
Yes, exactly, we’ve been doing this already, and this is just one more step along the way!