As you read this article, do you wonder whether it was really written by me or by a generative artificial intelligence (AI) engine like ChatGPT, Bing or Bard? While this is a novel question for the time being, very soon the dynamic of living with computer generated content will be normalized — whatever that means. Like the advent of the internet or the smartphone, generative AI and large language models (LLMs) will significantly change the way consumers live and expect technology to work.
How will chatbots disrupt corporate learning and in what ways will chatbots work to help reinforce corporate learning? There are things that it can do immediately and there will be other innovations down the road that will be even more useful. And, as with most new innovations, there are some pitfalls and disadvantages that corporate learning teams need to navigate around.
Immediate Benefits
Perhaps the most pronounced feature that LLMs have introduced is a huge leap in natural language processing (NLP) ability. Most of us are all too familiar with the limitations of Amazon’s Alexa and Apple’s Siri in being able to understand complex user intent. Sure, Alexa has no problem telling us what the weather is today, but asking Alexa how to handle a difficult conversation with a loved one won’t work. LLMs like ChatGPT can not only understand complex prompts like this but also can carry on a conversation about the topics, taking in feedback if we have clarifying questions or need a more nuanced answer.
This ability can help someone in the workplace that is stuck and needs some information or help getting unstuck. And rather than having to ask someone for help, someone can just ask a chatbot and get an immediate answer.
Because chatbots can carry on an ongoing conversation, you can now have learners practice communication skills with chatbots. For example, let’s say you’re a manager and you have a team member who is consistently showing up late to work. You can tell the chatbot to pretend to be this team member and to role-play with you. You can practice how you might approach this tricky conversation until you feel comfortable that you are ready for any and all possible outcomes. In the past, creating simulations like this would take a considerable amount of time to prepare but now, it is available right out of the box!
Much of the “wow” factor of these chatbots has been the ability for them to create content. In corporate learning, this can be a powerful tool to help rapidly (and inexpensively) create training content to reinforce your core training programs and people development initiatives.
Potential Pitfalls
The power of these new chatbots certainly has the world enamored (and perhaps terrified, too) but this whizbang technology is not without its drawbacks.
Your workforce may look to these chatbots for definitive answers to help them do their jobs — but you have little to no control over how these chatbot answers. One of the biggest criticisms of these chatbots so far is how inaccurate they can be from time to time. There are ways to seed the chatbot with proprietary content to help steer them to more contextualized answers, but you really don’t have the type of control that you may need to ensure absolute correct responses.
Many organizations need to worry about privacy and security about training content as well as personal information about its workforce. These chatbots are public services and so organizations need to be careful about what information their employees are feeding into these chatbot engines as well as what information do these chatbots capture about your users.
Usage of these chatbots is very much a personalized experience and so by nature, it is self-paced, informal and unstructured. Many organizations will demand some level of learning oversight to track and monitor what their workforce is learning and working on. So, these chatbots may be best served as a part of a broader technology platform to create some needed monitoring, structure and corporate reporting.
Down the Road
It’s amazing to see how much the first few months of these LLM-based chatbots have created in terms of market disruption. What might we see in the months and years ahead? Most certainly, we’ll see tools created to help organizations address some of the pitfalls listed above but we can also expect to see tools to help these chatbots do more than just answering user questions. Expect to see chatbots evolve to be able to proactively coach and remind and hold learners accountable over time.
Also, while these chatbots remain a one-on-one conversation for now, expect to see these chatbots be able to manage multiple relationships and groups at the same time. All in all, this chatbot technology will do much to help facilitate personal and professional development — which we can, and should, be excited about.
