Learning Agents Part 2: Learning Agents, Done Well

Posted on by Judy Unrein

The best example I’ve seen of a learning agent in action (Course Demo — free account creation required):

learningagent2

The learning agent from Allen Interactions' anti-terrorism course

The course teaches police officers and dispatchers how to identify and respond to terrorism threats. The initial content presentation is made by a learning agent.

Why I think it’s effective:

1) The agent’s appearance and voice are spot-on authentic; I suspect the developers used a real cop. Good call… I couldn’t see a namby-pamby voice talent impressing this audience.

2) Images and narration are used, which are every bit as effective as video, without the hefty bandwidth requirement. Also, making changes will be much easier than with video.

3) Since it feels like we’re in a classroom, the designers gave the agent good presentation skills. It’s not super-thrilling, but he’s not merely reading his slides to us.

Your thoughts? More examples?

Stay tuned: Learning Agents, Done Poorly

  • Harold Cypress

    On the use of “images and narration” rather than video to save bandwith, etc. – yes, very effective – the first time I used this mode in an asynch course I had 4 different agents/characters used in the course. The learners recalled the course after 1 day – 1 week as actually having true video. This mode also allowed us to choose the images for their impact and quality and then independently select the audio narrator. Then of course we could test and select from a range of voice actors, tone, style, etc. We also knew that we had high flexibility to record audio in several languages as well.

  • Sara

    Pretty nice post. I just stumbled upon your site and wanted to say
    that I have really enjoyed reading your posts. Any way
    I’ll be subscribing to your blog and I hope you post again soon!

  • Judy Unrein

    Harold — Great thought on different languages! That usually isn’t a consideration in my work (right now) but it’s a great one for many IDs out there. Can’t emphasize enough how much easier images/audio are to edit, as well, as opposed to video. If I have to change one sentence in a course, think about the ramifications of doing that for audio alone vs video…

    Sara — Thanks! I will be continuing this series all week and I’ve got lots of ideas for continuing… and if you have more, feel free to leave me a note in the Inbox.

    Sorry for being slow to moderate comments today, folks… I’m actually on vacation!

  • http://mobilemind.net Tom King

    These are characters. I’d like to see intelligent agents. Something the designer/instructional designer can use to monitor progress, adapt instruction, etc. Something even a student can use to delegate tasks or responsibilities, grant authority/permissions (or deny some) to act on one’s behalf and in one’s best interest.

    Now THAT would be truly dramatic and innovative for elearning.

  • Judy Unrein

    Tom — I went around a bit before actually using the term “learning agent” to refer to these examples, as I don’t feel that it’s a term that’s well defined by our industry. I think of characters as onscreen personas that are not in a “teaching role”; however, when it comes down to it I wouldn’t quibble the terms with you and I applaud you for valuing pushing e-learning past our comfortable status quo.

    I do agree that this example is not interactive, and on-screen personas can be interactive regardless of what you choose to call them or what role they’re in. My next post in this series (Part 4) will actually be about moving things a little more in that direction…

  • Pingback: Learning Agents Part 4: Another Agent to Learn From | onehundredfortywords