I am fascinated by how good the ChatGPT functionality is at generating written text. I have asked it to create blog posts and poems on a variety of topics, and it generates convincing output. Even news outlets are using it to generate content. It has me wondering whether the days of writing blog posts are over. Does the world need another blog on technical leadership?

A tool like ChatGPT seems well suited to digest a large amount of information (the Internet) and synthesize answers based on the trained information. I can ask “write a blog post on running an effective meeting” and I will get an organized output with some commonly held practices including “have an agenda” and “start and end on time.” I find myself thinking about two key missing elements: curation and innovation.

How does a tool like ChatGPT curate the information in the training set? AI does not have the professional experience to evaluate good advice and separate it from bad advice. The tool relies on the information provided for both facts and judgment. You can find almost an infinite number of recommendations for leaders. How does an AI tool differentiate good and bad advice?

I am reminded of a saying that “conventional wisdom” is not necessarily correct or best practice. The fact that most people do something a certain way does not make it a good or wise practice. On the contrary, I would say most managers are not “great managers” (think how many “great” managers you have had). How do you curate the advice you follow to become a better leader or manager? You curate the advice you take that provides you guidance to become what you experience as a great manager.

How does a tool like ChatGPT innovate beyond existing guidance? The tool is trained on and limited to a finite data set. It is then limited to either reiterating existing data from that data set (or creating something fictional). There is no real-world innovation or gains in knowledge from experimentation. ChatGPT is not leading a team nor is able to gain real-world feedback and therefore relies on those of us to do the work and share the outcomes for it to include in the training set.

As much as we may think management and leadership are well-covered domains, we must be open to the fact that change is the only constant. Imagine taking leadership lessons from an AI tool trained from before 2019 and applying leadership lessons to problems faced during the pandemic and resulting shutdown.

The lessons learned during the pandemic about what tactics and tools worked for leaders during a “black swan event” added something new to the data set. Remote work was required, no longer a perk, for many office workers. Empathy and flexibility became critical for everyone managing new outside stresses and potential mental health issues. The landscape for managing white-collar office workers was very different.

I have no doubt that tools like ChatGPT will become useful in the same way grammar checking or language translation have already become. Avoid aspiring to be an average manager. I recommend curating the guidance you follow and innovating to extend best practices in leadership.