4 Caveats Every L&D Professional Must Consider When Using Chat GPT

What To Look Out For When Using Chat GPT

As Learning and Development (L&D) teams are becoming indispensable advisors to CEOs and C-suite leaders in most organizations, L&D professionals are increasingly exploring innovative tools and technologies to increase the quality and speed of delivering and curating workforce learning experiences. One such tool that has gained considerable attention is Chat GPT, a form of generative AI and a powerful language model capable of generating human-like text. According to Gartner, Chapt GPT can be used to generate and improve prose and code development, summarize text, classify content, answer prompts, and translate and convert languages, including programming languages.

L&D professionals can use Chat GPT in many ways, including generating content and lesson plans in seconds. Additionally, L&D teams can use Chat GPT [1] for personalizing, tagging, translating, and search indexing learning content in a matter of seconds. While Chat GPT holds great potential and offers several use case opportunities in revolutionizing L&D, it also carries risks and biases [2]. As such, it is essential for professionals to use Chat GPT and similar generative AI tools with caution. This article will explore four caveats that every L&D professional must consider when incorporating Chat GPT into their learning strategies.

4 Things L&D Pros Should Consider When Using Chat GPT

Accuracy And Reliability

While Chat GPT can generate coherent and contextually relevant responses in seconds, it’s often limited regarding accuracy and reliability. It is crucial to remember that Chat GPT relies on scouring vast amounts of pre-existing data and patterns it has learned during the training of its model. According to Gartner, Chat GPT is not trained using data generated after 2021, and because it cannot cite its sources, it is known to generate fake academic articles that look real. Additionally, Chat GPT is not good at math. As a result, the accuracy and reliability of the information provided by Chat GPT may vary. This means that L&D professionals must exercise caution when relying on Chat GPT as a primary source of information or as an expert in a specific domain.

Instead, we strongly recommend that L&D professionals use Chat GPT as a brainstorming partner and supplementary tool that can provide insights and suggestions swiftly, and always verify and validate the Chat GPT results through other credible and reliable sources. Specifically, to mitigate potential inaccuracies, L&D professionals should consider implementing a system of fact-checking and human oversight. For example, when Instructional Designers develop new learning content, it is advisable to engage subject matter experts to review and validate the content generated by Chat GPT to ensure its accuracy and reliability. By leveraging Chat GPT as a support tool rather than a sole authority, L&D professionals can harness its speed and potential while maintaining the integrity of the learning experience.

Ethics

As Chat GPT becomes more prevalent in L&D, it is also crucial to address the ethical implications associated with its usage. According to research by the American Psychological Association (APA) [3], there are several ethical considerations L&D professionals must consider when using Chat GPT. Chat GPT learns from vast amounts of data, including internet sources, which may contain biased or discriminatory content. L&D professionals must be vigilant in ensuring that the language model does not perpetuate or reinforce any biases, stereotypes, or discriminatory practices. Additionally, Chat GPT can generate false information, such as fake learning theories, and list fake academic articles to support them. On the learning side of the equation, Chat GPT may prompt learners to cheat on formative or summative assessments.

To address these ethical concerns, L&D professionals should prioritize the inclusion of diverse voices and perspectives and check for cognitive biases during the development, refinement, and curation of learning content. This approach can help mitigate the risk of inadvertently perpetuating biases and relying on them to make assertions. L&D professionals are also strongly advised to engage their team to conduct regular content checks and evaluations, in order to identify and rectify any biased or potentially harmful responses generated by Chat GPT. It is easy to miss a bias if you are the only person looking at it. Additionally, while governance and ethics guardrails are being developed at the country and social level, in order to facilitate oversight, the L&D team can define and communicate a set of guidelines for your organization. This will ensure that your team is aware of these guidelines around language to mitigate ethical biases and limitations of generative AI, such as Chat GPT. By actively monitoring and addressing ethical considerations, L&D professionals can ensure a more inclusive and equitable learning environment.

Data Security

Another crucial caveat to consider when utilizing Chat GPT is data privacy and security [4]. Chat GPT operates on an open-source platform, so all data generated is available and searchable in the public domain. Chat GPT processes and stores user interactions, which may include sensitive information and Personally Identifiable Information (PII). L&D professionals must take appropriate measures to safeguard the privacy and security of organizational and learner data.

When using Chat GPT, L&D professionals should ensure that the content does not include PII, that learners’ consents are collected, and that they are using organization guidelines regarding data usage, storage, and retention. Additionally, L&D teams must collaborate with their IT partners to ensure that data encryption and secure storage practices are implemented to protect user information from unauthorized access. By prioritizing data privacy and security, L&D professionals can maintain learner trust and confidence while using Chat GPT.

Conclusion

While Chat GPT holds tremendous opportunities to transform and speed up the design, curation, development, and personalization of learning content, L&D professionals must remain cautiously optimistic when using generative AI forms, such as Chat GPT. Considering the four caveats discussed in this article, accuracy, reliability, ethics, and data security can help L&D professionals leverage Chat GPT effectively while mitigating possible risks and limitations. By complementing the power of generative AI technology with human critical thinking, expertise, and oversight, L&D professionals can design and deliver dynamic and engaging learning experiences that harness the benefits of Chat GPT and similar generative AI capabilities, while upholding the highest standards of quality and integrity.

References:

[1] The promise of generative AI and ChatGPT for L&D

[2] Chowdhury, Naem & Rahman, Sagedur. (2023). A brief review of ChatGPT: Limitations, Challenges, and Ethical-Social Implications. 10.5281/zenodo.7629888.

[3] How to use ChatGPT as a learning tool

[4] ChatGPT Has a Big Privacy Problem