From SAMR to CAFE: a New Integration Model for AI Technologies in Education

Whether I was trying to plan the work I am doing with AI at International School of Panama or to conceptualize it in writings, trying to use the traditional SAMR model to frame the educational integration of AI technologies proved unsuccessful. As many times as I tried, something was always missing. Or rather, many things–and some of the most important ones. This was true from the start, and is proving more and more obvious as new AI tools and capabilities are being developed every day. This impasse prompted me to develop the CAFE model, which can be aligned with SAMR, but is more refined and better adapted to artificially intelligent educational technologies.

---

The main benefit of its creation was to force me to reflect on the reason why AI integration did not fit into the usual SAMR spectrum - and thus on the difference between AI and previous edtech developments. This led to several insights that helped improve my understanding of these technologies and translated into many practical benefits. Indeed, CAFE does not only help educators use AI tools efficiently, effectively, and transformatively, but also respond to this ongoing revolution with agility, responsibility, appropriate focus, and a long-term orientation–thus ensuring that they channel its full potential, all while mitigating its threats.

Channeling

Looking at the visual representation of the two models, the color coding makes it clear that the CAFE model starts with an idea that is not present in its SAMR predecessor: Channeling. This first step has to do with the necessity to respond to AI developments by taking into consideration its risks and the updated use and ethical guidelines it calls for, as well as its disruptive power and the new skills it requires educators and future generations to acquire. This highlights the fact that AI technologies are different from previous edtech developments in that they are considerably more powerful and transformative, to the point that they require us to adapt and respond to their consequences, which reach much broader, deeper, and further.
Previous educational technologies were an opportunity. With them, the question was: how can we integrate technology into the school context? AI, however, is not a simple opportunity. It is a challenge. Students will use AI, which will transform many aspects of society– whether we “integrate” it or not. The question thus shifts from “how can we use this most effectively?” (SAMR) to “how do we have to respond to it intelligently and humanly?” (CAFE).

Automation

Next, the “Substitution” level in SAMR becomes “Automation” in CAFE. The two partially overlap, as AI can by definition complete some of the same tasks as humans, just more efficiently. There is still a key difference, however, as AI does not simply allow teachers and students to do certain things faster or more easily, but has the ability to do it for them. Here, it is not a previous tool that is being replaced, but the doing (task), and indeed the “doer” itself.

Focus

In the CAFE model, “Augmentation” and “Modification” are combined into a “Focus” stage. Just like previous technological advances allowing us to do more, or better, AI comes with many more options, and much faster, than educators can realistically put to good use. This bottleneck leads to important questions that, given its frantic pace, are even more relevant for AI than they were for previous edtech developments: how to overcome both choice anxiety and superficial novelty chasing? How to select the most promising tools and pedagogically sound uses, with the greatest potential benefit for students? In addition, because it can automate lower-order cognitive processes and assist more advanced ones, AI actually calls for an even tighter focus–on higher-order thinking and uniquely human capabilities.

Experimentation

Finally, the “Redefinition” level of SAMR is itself redefined as “Experimentation” in CAFE. Both address ways in which technology expands the horizon of learning and opens new, previously inaccessible or even unimaginable paths forward. “Experimentation” introduces three additional elements, however: a research-and-development “trialing” approach, a continuous evaluation of success and consequences, and a continuous refinement – all of which are due to the fact that AI technologies are both a historical turning point and an ongoing  revolution. 

An Agile and Adaptive Model

While it is still being debated whether the SAMR framework should be thought of as a spectrum or a ladder, both are equally linear. CAFE, however, is a circular, or rather an “agile” model – a fashionable expression that is often used quite loosely, but strictly means that the different phases of this process run in parallel and feed into each other, allowing to adapt adaptations as they are being implemented.

The hope is that the CAFE framework will help schools develop truly intelligent responses to the monumental threats and promises of AI for learners and educators, from the thoughtful integration of these technologies in the classroom to encompassing, bold, and responsible strategies at the whole-school level.