Weekly Newsletter

Practical AI Strategies

This week on the blog... A Taxonomy of Agentic AI

Hi everyone,

Happy Easter to those of you on a break. Term 1 has wrapped up here in Victoria, and I hope wherever you are in the world you get at least a few days to switch off. If you do find yourself with a bit of time over the holidays, this week's three articles might be worth a read with a coffee or two, because they cover a lot of ground: from the language of AI agents, to the hidden cost of AI-generated content, to what it means for students to take responsibility for their learning when a chatbot is happy to take it for them.

The question running through all three posts is: who is doing the thinking?

A Taxonomy of Agentic AI

"Agent" has become one of those buzzwords that means whatever the person selling it wants it to mean. Microsoft's version is different from OpenAI's, which is different from Anthropic's, which is different from what most people imagine when they hear "autonomous AI." In this article I provide a clear definition and a five-level taxonomy, from code-using chatbots at the bottom through to agent teams and swarms at the top.

Each level inherits the capabilities of the ones below it, and I walk through practical examples of what each level can actually do, including in education.

Read: A Taxonomy of Agentic AI

The Effort Economy of Slop

I came across a definition of AI slop that crystallised something I've been thinking about for a while: slop is something that takes more effort to consume than it took to produce. This article explores the effort economy of communication, and how GenAI inverts the traditional relationship between producer and consumer.

When a student submits AI-generated work, the teacher becomes the primary meaning-maker, doing all of the cognitive work the student declined to do. The same dynamic plays out across emails, policies, newsletters, and resources. If you've ever opened a document and thought, "a human didn't write this," you've felt the inversion of effort.

Read: The Effort Economy of Slop

Gradually Reclaiming Responsibility

The gradual release of responsibility model is one of the most widely used frameworks in education: I do, we do, you do. But when AI enters the picture, there's a risk that responsibility is released not onto the student but onto the chatbot.

This article blends the GRR model with the concept of "resistance" from last week's article, mapping the increasing effort required to maintain ownership of thinking as AI use deepens. In the original GRR model, responsibility flows from teacher to student. With AI, there's a third party in the room, and it's perfectly happy to take on all of the responsibility without ever intending to give it back.

Read: Gradually Reclaiming Responsibility

Cheers,

Leon

PS: All courses and digital downloads at Practical AI Strategies are 25% off over the break. Use the code EASTER-2026 at checkout. Ends April 26th.

---

Stay informed.​
The Practical AI Library curates fresh, practical articles, research, and advice on AI and education—updated every week. Check it out
Check it out → Practical AI Library​

---