Why AI and “one right answer” don’t get users to their end goals
- Presentation
- Artificial Intelligence (AI) in Technical Communication
- 05. November
- 16:30 - 17:15 PM (CET)
- Plenum 1
- finished
- No presentation slides uploaded
-
Joe Pairman
- Tridion (RWS)
Contents
Delivery of tech docs often focuses on search — getting users to a good starting place. But users often have goals that a single page can’t cover. Serious tasks can require reference info, scenarios, and guidance. Users may need information in different media. Each user’s goal depends on their specific context, and they need to follow a unique path through the information. Writers cannot possibly design content for all of those users with all of their unique paths.
Many of the attempted techncal solutions in the field have been at best a partial success. At various times, tech docs have relied more on search, or more on manual hyperlinks, or taxonomy. Pages have been long or short, and other media have come and gone. Now AI offers easier ways to map content together, but even that’s not enough. Why?
Only a combination of techniques will give you the information model that you need. And the techniques on their own won’t get you all the way — you need an integrated design to bring your solution together. Learn how in this talk.
Takeaways
- What AI can do to suggest appropriate content — and what it can’t.
- The real needs of enabling complex user tasks, and how to map them to usable metadata.
- How to design a suitable goal-enabling solution for your own organization’s needs,
Prior knowledge
The audience should have some experience with online content delivery.
It may also be helpful (but is not necessary) to have some experience with search applications, or taxonomy/ontology, or natural language processing and AI.