AI for course creation? Let the trainer beware!

Artificial Intelligence (AI) technology applications are numerous and increasingly tantalizing for those of us looking to save time and money in creating and editing training content. While AI is proven to produce text quickly and efficiently, the risks and challenges are real and merit oversight.

1. Accuracy

   Even the advanced language models are not infallible. They can talk a good game but do not inherently understand content in the same way humans do. Lack of deep comprehension can lead to inaccuracies or misleading information. If the AI model is trained on biased and/or outdated data, these errors will/can be perpetuated and result in incorrect or even harmful information.

   Consider this news release from 2023:

   CNET AI plagiarism/content controversy
   "Technology publication CNET finds itself in hot water after quietly publishing a range of AI-generated articles that include factual errors and cases of apparent plagiarism. The technology publication ends up issuing corrections relating to 41 of the 77 news stories... AI content generation has created a new set of significant ethical and financial considerations for websites across the digital publishing industry, as well as educational institutions like universities and schools. CNET is unlikely to be the last company hitting the headlines for this kind of AI mistake."

2. Ethics

   Fairness and inclusivity are essential to training material. AI, again drawing from outdated data, can reinforce gender stereotypes as well as exclude certain demographic groups. Human oversight is essential to avoiding harmful biases and in ensuring ethically sound material.

3. Relevance

   Up-to-date training material constantly considers industry standards, new rules and regulations, learners' background and evolving trends. AI typically struggles to adapt content to the most current information and other contextual subtleties.

4. Legally compliant

   Training on regulations requires that material is legally compliant and necessitates human oversight. You cannot afford to train on outdated information.

Bottom line: Balance any use of AI in creating or editing training material with a Subject Matter Expert (SME) who can attest to the accuracy and fairness of the content. 'AI did it' won't win in court.

P.S. Worth a read from the New York Times August 26, 2024: 'When A.I.'s Output Is a Threat to A.I. Itself.'