Co-Intelligence, Post-Plagiarism and Arizona's AI Guidance for Schools

Co-Intelligence, Post-Plagiarism, HITL, Inequity, Continuous PD, Instructional Redesign?

May 12, 2024

Last week, Arizona released its AI Guidance for Arizona Schools. As my students debated this topic over the weekend, I had a great excuse to dig right in.

I must say that I’m a bit biased because it is awesome to have one of my blog posts referenced in the document, but I also must honestly say that this is a strong guidance document. It not only reinforces some important ideas that have been shared elsewhere but it references so many other critical ideas that I don’t think have yet been shared in K-12 guidance documents.

I do wish it covered instructional redesign. The grammar of 20th-century learning is dead and instructional redesign needs prioritized.

Newly Shared Ideas

Co-Intelligence

While many hold to the idea that we will always be smarter AIs and/or that we have some “unique” human qualities, I’ve become more and more skeptical of these claims. AIs are already smarter than us in some narrow areas such as general knowledge, predictions, and image classification. AIs are starting to demonstrate creativity, empathy, and critical reasoning abilities. Applications such as hume.ai can even anticipate when we will laugh before we laugh and laugh with us (that may make some of you cry).

I don’t think trying to compete with AI is really the path forward. The best path forward, I think, is to take our own intelligence and extend it with AI. Scarier terms for this are “cyborg intelligence” and ‘“augmented intelligence,” but I think “co-intelligence” is the most acceptable. Dr. Chad Gestson, Executive Director Arizona Institute for Education and the Economy, writes in a forward to the document.

In practice, I think it’s sort of simple —

How do students use AI to be better debaters? To do better on a test? To do a better presentation? To produce a stronger portfolio? How do we all use it to teach better? To write better? To be better parents?

Post-Plagiarism

This is the first K-12 guidance document I’ve seen explicitly tackle the idea that we need to redefine what constitutes plagiarism (NC’s newest version does introduce this idea).

Although they don’t explicitly cite Sara Jane Eaton’s work on post-plagiarism on this page, they do in the reference section.

It is going to take education a bit of time to accept this reality, especially given that people still fantasize that normal writing assignments can continue with writing detectors stopping the “cheaters,” but this debate is old and done.

My Normal “Human in the Loop” Concern

Like all the other guidance documents, it makes a strong case for protecting the role of the “human in the loop,” which refers to the idea that the human should be the final decision-maker.

It’s conceptually difficult to argue against this but as I noted at the beginning, AI is already smarter than us in some narrow ways. In education, there is already some evidence it is better at grading essays than humans on a couple of common metrics. It will almost certainly be more consistent, especially across teachers, buildings, districts, and states. Parents and their attorneys may start using AIs to challenge IEPs, course placements, and disciplinary decisions.

Generally, yes, I support HITL, but as AI gets smarter than us in more ways, this is going to get complicated fast.

Common Areas

This document hits on common themes, but in some cases adds a lot more detail (and nice formatting) that helps.

Prioritization of Job Training

This document emphasizes the role of education in promoting job training and references the most recent work on the need to reskill.

Of course, what it leaves unanswered as to what jobs we need to upskill for. That’s not a criticism, as no one really knows, and one could even make the case that AI will just be doing all the jobs, leaving nothing to upskill for (Cuban). The only job the World Economic Forum document cited in this report argues we should train students for is machine learning scientist.

Strong Stance on the Digital Divide

The document takes a strong stance on how failure to not only provide access but also training on use will strengthen the digital divide. Proper training for both teachers and students is essential for students to make sure everyone uses this in a way that captures the benefits

Closing this gap requires action by schools. If schools don’t act to promote equitable AI training and use, the gap will only widen, and this time it will be with one set of people using AI to amplify their intelligence and the others not using it.

Every day public schools wait to start thinking about how to prepare their students for an AI World, inequality grows.

Writing Detectors

The document repeats the common criticism of writing detectors that can also be found in a couple of other guidance documents.

What I really like about this document is that it emphasizes a point I’ve been trying to make -- students who are AI literate, which tend to be higher SES students, know how to beat the detectors. It’s not the case that if you run a paper through a detector and it gets a low percentage score it means it wasn’t written heavily with AI. It may have passed out of luck or because a student knew the techniques to beat the detector. Who gets caught? Struggling learners who tend to be lower SES and/or ELL.

This writing detector business is as admirable as selling cigarettes. Imagine if the money schools are spending on these things was redirected to preparing students for the AI World.

Commitment to AI Literacy

The document not only makes a commitment to AI literacy but breaks it down in a way that I think is helpful.

Additional Areas

The document covers many ideas found in other guidance documents that we have been emphasizing, but I want to list a few here with noted points of emphasis.

“Continuous” professional development.
“You DO have staff members or community members with thoughts. Call them together for a facilitated conversation about the direction of AI implementation and create your task force from that group, having already conducted your first needs assessment. “
“Stay vigilant for transformative changes to teaching and learning methodologies and congruence with system values.”
“GenAI can perform some administrative and organizational tasks better than a human can. Consider its ability to analyze student data quickly and accurately. GenAI can also help streamline administrative tasks, freeing up time to focus on more impactful activities. “
“GenAI offers capabilities that can allow students to be more independent and self-directed in their learning journey. Students may find additional uses that bring out curiosity, creativity, and reflection. “
”GenAI offers teachers a wide range of applications that can simplify routine tasks, create learning materials, or serve as a thought partner.

What’s Missing? Instructional Redesign

“We can challenge the traditional notion of cheating and plagiarism by redesigning curriculum and assignments that students are eager to learn and are willing to invest time and effort into. Assignments that encourage authentic student engagement and creation can rarely be completed via automated solutions.”

This is a quote from the “Redefining Academic Integrity” part of the document that I referenced earlier.

If you are a teacher or anyone with any experience in education, you are thinking, yeah, sure, sounds great, but does anyone have any ideas? Isn’t this the Holy Grail of education? Isn’t this what we are all mad at the social media companies for figuring out at our expense?

This is the hard part. AI literacy, media literacy, professional development, App evaluation, policy review, etc., all take time and money, but they are all quite simple compared to the “instructional redesign” (Crisci, Furze) needed for a world where the future of employment is uncertain at best, and the development of durable and metacognitive skills are more important than ever.

But it’s not that hard.

This is what you need to do.

Support student debates both in and out of the classroom. This will support metacognition and the development of durable skills, which are already the number one job skills. We wrote a whole paper on it. I have lots of expertise in this area and I have friends. We can help you infuse debates into all of your academic programming.

Gamify Instruction. My friend Erick Tucker wrote about this today.

Develop entrepreneurship programs. The number of jobs will decline, at least in the short term. Some think they will never recover, as we can simply have the AIs do what employees would otherwise do in a new business.

And, no, jobs are not inevitable. Look at societies around the world that have 20-30% unemployment. AI may produce a restructuring of society along similar lines. And we’ve never had a world where hype intelligence can be bought and sold for pennies on the dollar or the price of a “meal deal” at McDonald’s.

What does this mean for education? Well, it means students need those durable and metacognitive skills, but it also means they need to learn how to start their own businesses and manage their own teams of AIs (LeCun).

And it sure won’t hurt to learn how to advocate for themselves.

Support project-based learning, interdisciplinary learning, and phenomenological learning. Think about what you remember from high school (I mean what you remember of the academic stuff, not the other stuff); it was probably stuff that involved these learning strategies.

Help students develop portfolios that demonstrate their capabilities (Quidwai). Demonstrating capabilities is the current focus of AI developers (Suleyman).

I’m sure there are other ideas. Many educators spent 25+ years being educated. I’m sure they can figure this out.

Modern assessments need a retirement party (Bickerstaff).

The grammar of 21st-century learning is dead (me).

There is no value to educational methods that prepare students for a world that does not exist (me).

As noted in the Santa Ana guidance (Quidwai, Almandarez) and the NC guidance, instructional redesign is essential.

Santa Ana