Similar to when Google first became integrated into our daily work life, it will take time for people to become familiar with ChatGPT’s capabilities and how to use it effectively. I believe that concerns about ChatGPT making formal learning obsolete are unfounded. The picture is more nuanced and so it helps to be clear about which use cases play to ChatGPT’s and formal learning’s relative strengths.
Where does ChatGPT win?
Speed: L&D’s lead time for formal learning is expressed in weeks or months, whereas ChatGPT provides almost instant results. Interestingly ChatGPT will help reduce L&D’s lead time as it will reduce the time commitment needed by subject matter experts and accelerate several writing tasks.
Generic content e.g. Excel formulae, introductions to management concepts: with well-written prompts, ChatGPT returns almost instant deliverables e.g. tailored instructions, summaries, briefing papers, backgrounders. However, if context is not important to the learning, you probably shouldn’t have had a bespoke course anyway (the need was already covered by Google or your LXP).
The long tail of learning needs: individuals can get answers (hopefully good enough) to a vast range of prompts. L&D is resourced only to use formal learning to solve for strategic learning needs relevant to a critical mass of people.
Where does formal learning win?
Learning experience: formal learning is more that information, it should be engaging, varied and memorable. ChatGPT’s user experience is basic, the clever part is behind the scenes.
Tailored learning: effective performance consulting should lead to a clear and well communicated context for a personalised, formal learning experience and call to action. ChatGPT returns the “average of the internet” for a given prompt.
Reliability: L&D should go through the process of validating content with the organisation’s subject matter experts so that it can be relied on. ChatGPT’s algorithm surfaces content that is limited by the data model e.g. may contain mistakes and not include recent developments. There is no trail to allow checking for reliability.
Relevance: L&D review and refine formal learning content so that it ties in with the organisation’s purpose, strategy, culture, brand and tone of voice. ChatGPT’s algorithm estimates what is relevant or not and is subject to limitations in the data model.
Authenticity: messages and experiences are based on real people working in the organisation, whereas ChatGPT returns the “average of the internet” for a given prompt.
Confidentiality: there is minimal (or at least well understood) risk of leakage of questions and answers beyond the organisation from a formal learning course. There is a risk that ChatGPT users input confidential/commercial information that they should n’t and this gets leaked through incorporation into the language model.
Reward: a user can be given credit/qualifications/CPD points for attending/completing formal learning as well as any performance system reward for performance improvement. As learning is not visible in ChatGPT, although the user may get rewarded for performance they won’t get rewarded for learning.
Conclusion
Undoubtedly, individuals will utilize ChatGPT to solve certain problems for which they would have previously submitted a learning request to L&D. However, L&D can still successfully differentiate the features and benefits of formal learning (courses and e-learning). With effective internal marketing and communication, L&D can establish a strong, defensible and well understood position for its strategically important formal learning offerings.