Search results

Links from around the web -- June 10, 2024

by Tom Johnson on Jun 9, 2024
categories: technical-writing

The following are interesting reads or listens related to tech comm. The summary bullets are AI-generated, followed by my own non-AI-generated thoughts. Topics include podcasts on RAG techniques for AI content development, OpenAPI reference guides, dead-end counterarguments, Lavacon in Portland, and AI cautiousness.
  • Retrieval Augmented Generation (RAG) improves large language model output by retrieving relevant information to supplement the model’s knowledge.
  • Technical communicators should optimize content for AI by cleaning, chunking, and adding context and metadata, partnering with engineers building RAG systems.
  • RAG and AI-assisted tools can accelerate content creation, but human writers remain critical for refinement, quality, and style adherence.

Tom’s thoughts: RAG works — it’s the secret to making AI tools generate more accurate, usable content. This podcast does a nice job explaining the approach while also acknowledging that the first AI-assisted draft is still a long way away from being publication ready. See my post Gathering source material for context input for more details. (BTW, I appreciated the mention of my site in the podcast.)

  • Speakeasy has released a comprehensive, open source reference guide for writing OpenAPI specifications, with AI-search to quickly answer questions.
  • OpenAPI provides a standardized way to describe RESTful APIs, enabling greater visibility and consistency across API documentation, SDKs, and other surfaces.
  • The guide covers OpenAPI document basics, format and file structure, and prioritizes approachability and practicality for developers of any skill level using OpenAPI 3.0.x and 3.1.x.

Tom’s thoughts: This is an excellent tutorial for developing an OpenAPI specification. In many ways, it supplants the Step-by-step OpenAPI tutorial in my own API docs course. It will be interesting to see how AI transforms the coding and syntax wrangling with OpenAPI specification development. At the very least, this info could augment your AI chat as you use AI to help create each of these sections in your specification.

  • Technical writing is a versatile profession that can serve as a platform for transitioning into other roles within the tech industry, such as development, product management, or information architecture.
  • While the technical writing profession may evolve due to technological advancements and shifting priorities, its core essence of bridging the gap between technology and people will remain the same.
  • Technical writers seeking career changes should look for opportunities within their current company, while those seeking their first job should focus on selling their skills and adaptability rather than limiting themselves to specific job titles.

Tom’s thoughts: There are a lot of Reddit posts from people wondering if tech writing is a dying field now that AI can do so much. I’m surprised and relieved that even using AI tools as much as possible, replacement of my tech writing role seems unlikely. I still have a seemingly endless number of tech writing bugs to tackle, and documentation is highly valuable.

  • LavaCon 2024, held October 27-30 in Portland, Oregon, will emphasize the role of content in reducing costs, generating revenue, and improving customer experience.
  • The conference features presentations by well-known content strategy and technical communication experts, hands-on workshops, networking opportunities, and engaging evening activities.
  • Attendees can save $200 on in-person registration using the referral code “XMLPress,” while a virtual conference experience is available for those unable to attend in person.

Tom’s thoughts: It’s neat to see Lavacon take place in Portland this year. Last year it was in San Diego and before that in New Orleans. This conference has been consistent year over year, in part because Jack Molisani knows how to deliver a great conference experience.

  • Generative AI is best suited for productivity enhancement in content creation rather than generating content from scratch, helping to refine human-authored content.
  • Intellectual property issues remain a significant concern with AI, as the regulatory landscape is still evolving and varies across countries, leading some companies to establish their own AI usage guidelines.
  • The proliferation of AI-generated content has negatively impacted web search quality, potentially driving companies to adopt walled garden approaches to curate and provide reliable information to their users.

Tom’s thoughts: This podcast definitely pushes back against the AI hype, even dismissing AI as a tool for content creation. I think RAG techniques can overcome many of the hallucination issues, but I like their balanced and cautious perspective, especially noting the unresolved intellectual property issues.

About Tom Johnson

Tom Johnson

I'm an API technical writer based in the Seattle area. On this blog, I write about topics related to technical writing and communication — such as software documentation, API documentation, AI, information architecture, content strategy, writing processes, plain language, tech comm careers, and more. Check out my API documentation course if you're looking for more info about documenting APIs. Or see my posts on AI and AI course section for more on the latest in AI and tech comm.

If you're a technical writer and want to keep on top of the latest trends in the tech comm, be sure to subscribe to email updates below. You can also learn more about me or contact me. Finally, note that the opinions I express on my blog are my own points of view, not that of my employer.