Conceived in 2010 by education researcher Ruben Puentedura, the SAMR model has been a powerful conceptual tool when implementing technology in education.
Initially we operate at the lower end of this model and technology is used purely as a substitute to what previously was done often on paper. Although not a strict hierarchy, as we gain confidence, capability and functionality improves we can start to augment what we do and in time consider modifying and perhaps even redefining tasks and education in general.
I feel the implementation of AI is following a very similar model. Mark House from RM described AI taking us from a complicated situation to a complex one. In complicated systems (such as schools!) with sufficient knowledge and experience, outputs are generally predictable from inputs; with a complex one this is not the case. For example, with large language models nobody, not even the creators, can predict what the output is going to be for each prompt.
When we have such complex systems, the key is to take small steps - it is difficult to see where this complexity will take us in two to three years’ time. Therefore, it makes sense to move up the SAMR model slowly.
In general, we are currently using AI to substitute tasks which previously would have been relatively labour intensive. For example, AI can help with lesson planning, report writing and some assessment. However,ultimately, the approach to education is unaffected, just, hopefully, educators are more efficient (and less tired!). The exciting prospect, of course, is that AI will allow us to move up the SAMR model. Educators are already beginning to use AI to augment, i.e. improve, learning tasks. How long will it be before we use AI to modify significantly what we do and perhaps even completely redefine education?