I always leave conferences with one session in my head that stands out above the rest. I just returned from the Sloan C International Conference on Asynchronous Learning Networks and would cite a session on “Using Intelligent Agents” as that one session. In case you are unfamiliar with the term here is a link to the Wikipedia definition for intelligent agents.

Steve Knode of the University of Maryland University College was the presenter. His overlying premise was that vast quantities of information are easily findable now but the ability to effectively filter and fuse the information for specific users and circumstances makes for shallow knowledge and analysis. As intelligent agents, now in first generation, continue to develop, information will continue to increase in relevance and depth. The agents themselves will better learn and adapt to users and circumstances.

How will this further learning? I was running various scenarios in my head throughout the presentation as to how really effective bot development could change the learning landscape. The one futuristic scenario I keep seeing is that scores of highly intelligent and adaptive agents could be developed, each for specific content areas. Likewise, these bots could be mixed and match for hybrid topics or creating new topics. As long as there was a standard for interoperability, it would be easy for learning architects to add bots to specific challenges in learning modules, simulations or games as learning support. When a learner needed just-in-time information, it would be instantly available locally in the sim or module and completely “fused and filtered” per the users questions and circumstances. Best yet, the bot would itself be learning from the learners themselves and readjusting how it delivered its results.