Influence and Documentation in the age of Generative AI

Reflections on the perception and value of Documentation in the world of Generative AI.

In 2023 I spoke about how you - as an Individual Contributor in Tech - could “Scale Your Influence” through Documentation.

During that talk I quipped that AI probably wasn’t at the point where it could replace you directly, and presented the idea that the Documentation you create is an extension of yourself, your identity, and that its existence builds your influence within an organisation.

I don’t think this documentation-as-an-extension-of-yourself concept has fundamentally changed, but advancements in Generative AI really do challenge “the playbook” and balance of impact that writing documentation has when building and exerting influence.

One of the insights I’ve noticed with the increased adoption of Generative AI in our organisation is:

The simple existence of content matters less than it did pre-AI.

The ability to express an idea in written words is now much less valuable. Generative AI makes it easier for anyone to create tailored, substantial content with much less effort. As a Staff Engineer, if your influence and reputation comes as a result of your prodigious ability to create documents or write strategy, or just transfer your ideas into written form, that advantage you have is mostly redundant.

The optimist in me believes that this is a good thing - that giving more people more tools to express themselves on a more level playing field creates a greater chance of better outcomes, as a sort of merit-based democracy of ideas.

The pessimist says that the use of Generative AI (by anyone) to create Documentation devalues the inherent value of existing and new human-created Documentation. If everyone can create well-written, comprehensive documents and persuasive proposals with much less effort, how do we stop being overcrowded with possibilities?

From a practical point of view, for someone trying to establish and develop their reputation through Documentation, the quantitative approach is going to yield much less reward in the age of Generative AI.

Generative AI is likely to be just as good at writing empty and meaningless content as it is creating top-quality content, but instinctively understanding these differences is hard, for both the “author” and the eventual audience. Writing good and useful Documentation remains an art form and learning to write high quality content is still a valuable skill, but now the barrier to entry has dropped, there’s a shift in the perception of its value by the audience.