Yesterday, the World Economic Forum's “Global Risks Report 2024” ranked AI-derived misinformation as a top global risk ahead of climate change, war, and economic weakness. NPR reported this morning that Taiwan is struggling with misinformation as its presidential election happens this Saturday. According to CNBC, the United States has the NSA and FBI preparing to combat deepfakes and misinformation as the US prepares for the 2024 election.
The SEC X.com Account Was Hacked
The SEC (Security and Exchange Commission) urges people to be careful of misinformation and then, the unthinkable happened.
Careful what you read on the internet. The best source of information about the SEC is the SEC.— U.S. Securities and Exchange Commission (@SECGov) October 16, 2023
This week, according to American Banker, because the SEC had not enabled two-factor authentication, a hacker posted that the commission had approved the trading of bitcoin, resulting in this message on X.com.
The @SECGov X account was compromised, and an unauthorized post was posted. The SEC has not approved the listing and trading of spot bitcoin exchange-traded products.— U.S. Securities and Exchange Commission (@SECGov) January 9, 2024
We Can No Longer Tell What is AI-Generated or a Real Human
So, yesterday, as I entered the second day of teaching about AI and the “reality threshold” we have surpassed with AI image and video generation, truth was front and center. (See yesterday's lesson plan about how I start discussing AI.)
Step 1: What photo is real?
Using faces I generated from ThisPersonDoesNotExist.com, I asked students to vote on which face(s) were real. We discussed the reality threshold, making it very difficult for us to tell what is real and what is not. The answer is that they all were AI-generated.
Then, we headed over to thesecatsdonotexist.com.
Finally, we played “Which face is real,” and students admitted they struggled to tell what was real.
Step 2: Review Sports Illustrated's Real World Scandal Pretending AI Generated Content was Human Generated (and Introduction of AI Policy)
Then, I had them use perplexity.ai to see what they could find about Sports Illustrated's controversy with AI. I had them use this tool for several reasons. First, I want them to see that generative text with citations from AI can be very helpful when you have a piece of information like “Sports Illustrated had a controversy with AI, what was it?” Secondly, I wanted to introduce my AI policy in the classroom.
So, Sports Illustrated created fake reporters with stock photos of people, fake bylines, and AI-generated articles. When someone looked up the photo of a reporter and found it was a stock photo, they were “caught.” So, my students and I discussed that, and I shared a vital principle of AI.
Artificial Intelligence is designed to imitate human behavior. However, we should never allow AI-generated content or AI chatbots to “pretend” to be human. The origin of content should be transparently revealed. AI is not human. It is a tool. To pretend to be human is a lie. Students had a great discussion about this.
Then, we discussed the AI policy for the class, that if students use AI, they must cite both the AI conversation with a link and then the original source documents that the chat with AI led them to. If there are no source documents, students must find, verify, and link to them.
Step 3: What is AI's role in the learning process? (A discussion of PhotoMath)
Then, we moved on to discuss the need for two things with learning:
- Productive Struggle
- Focusing on the process of correctly learning how to do something
As we work to learn there is a certain amount of productive struggle required. (Read the Edutopia article I wrote about how to use AI to Encourage Productive Struggle in Math.) We cannot use AI to interrupt the struggle required, but it can be used to coach us on the learning process.
So, then we focused on math and discussed the unspeakable app in most schools — PhotoMath. (Lest you think I was telling kids about an app that they didn't know about – 100% of them admitted they knew what it was.) If you're not aware, PhotoMath is an app that lets you take a picture of the math problem and then it works the problem for the student. They did a think-pair-share, and their answers were pretty spot on.
Based on what I taught them, I asked what use of PhotoMath would be “cheating,” not only dishonest but also cheating themselves out of the learning experience. And what use of PhotoMath would enhance learning.
Dishonest Use of AI “cheating” at math:
Take a picture and copy the answer. Don't look to figure out what you did wrong, just copy.
Use of AI to help learning math:
Work a problem out completely. Then, check the answer with Photomath. Follow each individual step to see where there may have been a mistake and check the answer. If there is a mistake, use an AI tool to create a similar problem and work on that. Then, check the process and the answer using an AI tool. That would improve learning.
What do you think?
And yes, I agree with them. The second option respected both productive struggle and the process of correctly working the math problem. AI was used to coach, not to circumvent the learning process.
Step 4: Get Kids Ready to Talk
My first two days of discussing AI ethics are designed to get students thinking. To help them face the reality of AI and how it works and also the deceitful nature of it when used wrongly both to create content that is misrepresented as human-created or being a human being and also the role AI can play in learning.
Why AI isn't a dirty word
So, misinformation and disinformation are important topics, and perhaps one of our big issues in schools is treating AI like a dirty word instead of the learning tool it can be.
For example, I'm not the only one at my school embracing AI for learning.
I happen to teach at a school with an amazing English teacher who has taught the kids how to use AI as a peer review that offers suggestions based on comparing their papers to rubrics for getting feedback. Students link to the AI conversation that helped them improve their papers. That is awesome. We're seeing better writing results in this way. I have much to share from her and plan a future post on this topic.
But we must talk about AI. The kids are. We should talk about it with one another and with our students.
To be mute is to be silent on the major issue of our time. AI. Misinformation. They are inescapably intertwined. As I've said before: A Gullible Public is a National Security Issue. How can we call our students educated if we ignore one of the most important issues of our time.
Get 10 Habits of Bloggers That Win
I send free lessons with my newsletter. When you sign up, you'll get this ebook to help you understand blogging.
You'll get 3-4 emails a week.