This post was originally published on LinkedIn.

This may be a warm-ish take in GTM/MOps circles. But the era of sitting down for hours and writing up what you built, adding screenshots, linking to other docs, adding to a centralized knowledgebase maybe over.

Note: this is coming from someone who has spent years telling people to document more (sorry)

But just think about what we have now:

1. AI is very good at auditing metadata (i.e. how things are set up in your systems: fields, objects, automation, etc). We have seen 5-10+ hour audits of SFDC reduced to <30mins with Claude Code. This is real, not hype.
2. Most calls are recorded so we have call transcripts where the team talked through the decisions, the why behind things, sales training, etc
3. AI can search Slack messages on decisions, pivots, people getting alignment, etc
4. AI finding information in Google sheets/docs and other working docs showing decisions, discovery, mappings, definitions, etc

Technical documentation is just structured data/text from unstructured sources.

Before a human had to do it, but now AI can probably do it just as good.

Or maybe we don’t even need many written docs anymore, as any time someone asks a question AI can answer it for them based on your company data, metadata, and context.

Just ask a question. No more writing docs and no more having to read through a 25 page doc to find how to convert a lead.

And the less time the Ops team spends on writing up documentation the more time they have for building.

It likely a way better use of time right now to focus on how to train an agent to do this vs. writing up docs.