OpenAI’s November 23 announcements – what they mean for Technical Writers

We asked ChatGPT what OpenAI’s recent announcements might have on Technical Writers:

1. Enhanced language models
   • The new GPT-4 Turbo with a 128k context window can assist Technical Writers by allowing them to work with more extensive documentation in a single prompt.
   • Note: This is roughly 300 pages of content
   • This means you can reference larger portions of texts, such as entire manuals or multiple API documentation pages, without having to break down the information into smaller chunks.
2. Reduced pricing
   • The reduction in API prices could make it feasible to automate larger parts of the documentation process or use AI assistance more frequently.
3. Assistants API
   • The Assistants API can enable the creation of more sophisticated and specialised documentation tools.
   • Technical Writers could build custom AI-powered assistants tailored to their specific needs, such as extracting key points for summaries, or converting technical specs into user-friendly guides.
4. Multimodal capabilities
   • The integration of multimodal capabilities, including DALL·E 3 for image creation and text-to-speech, can help Technical Writers enrich their documents with relevant visuals and audio files, making the content more accessible and engaging.
5. Improved instruction following and formats
   • The improved ability of models like GPT-4 Turbo to follow instructions and support for formats like JSON can help Technical Writers generate structured outputs with less manual editing.
   • This can be particularly useful for creating code documentation, API responses, or data format samples.
6. Reproducible outputs and debugging
   • The new seed parameter enables more consistent/reproducible outputs.
   • This could help Technical Writers in maintaining consistency across document versions and enable better debugging of the documentation process.
7. Function calling updates:
   • With enhanced function calling, Technical Writers could more efficiently integrate external data sources or APIs into their documentation workflow, potentially automating parts of the data fetching and example generation processes.
8. GPT-3.5 Turbo improvements
   • The updates to GPT-3.5 Turbo, with a larger context window and improved format following, provide Technical Writers with a more robust tool for working on complex documentation tasks.
9. Persistent Threads in Assistants API
   • The ability to maintain infinitely long threads allows for more complex and ongoing interactions with the AI, which can help in building comprehensive documentation over time without worrying about context limitations.
10. Code Interpreter and “Retrieval”
    • Technical Writers can use the Code Interpreter to automatically generate code examples.
    • The Retrieval function can help pull in relevant information from large databases or documents, which might be useful for technical documentation involving code or data specifications.

See also: Using Generative AI in technical writing training course