Skip to content

Commit

Permalink
redirect genai 2023 outlook post
Browse files Browse the repository at this point in the history
  • Loading branch information
kaycebasques committed Feb 24, 2024
1 parent 31aaaf3 commit 48be26a
Showing 1 changed file with 3 additions and 126 deletions.
129 changes: 3 additions & 126 deletions src/content/blog/generative-ai.mdx
Original file line number Diff line number Diff line change
Expand Up @@ -10,129 +10,6 @@ description: >
documentation. This post summarizes what's going on.
---

<p>{frontmatter.description}</p>

## Potential threat

The threat is simple: [generative AI] has the potential to reduce the demand for
technical writers because it's getting pretty good at our job.

Take [ChatGPT] for example. You ask it questions in a familiar chat UI and it
responds with answers that have been synthesized from a variety of sources.
A lot of these answers are actually relevant and helpful. It can respond in
text, [code], and [image].

If it's any consolation, the generative AI community does not have a vendetta
against technical writers in particular. It is potentially a [general purpose
technology that may affect most jobs].

### ChatGPT's take

This is how ChatGPT (GPT-4) responded to the following question: "What parts of
technical writing are most likely to be automated by ChatGPT?"

> As an AI language model, ChatGPT can potentially automate several aspects of
> technical writing, making the process more efficient and less time-consuming
> for human writers. Some areas that are most likely to be automated include:
>
> Basic content generation: ChatGPT can generate paragraphs or sections based on
> given topics or outlines, providing a starting point for technical writers.
> This can speed up the content creation process and help maintain consistency
> in writing.
>
> Data analysis and interpretation: AI can analyze large datasets and generate
> summaries, trends, or insights that can be incorporated into technical documents.
>
> Formatting and template creation: AI can automatically apply formatting and
> styling rules to documents, ensuring they adhere to specific guidelines or
> templates.
>
> Grammar and spell-checking: ChatGPT can identify and correct grammatical
> errors, spelling mistakes, and other language inconsistencies, leading to
> higher-quality content.
>
> Terminology consistency: AI can help maintain the use of consistent
> terminology and phrases throughout a document, reducing confusion for readers.
>
> Content summarization: ChatGPT can create concise summaries or abstracts of
> longer, more complex documents, making them more accessible to a wider audience.
>
> Content translation: AI language models can translate technical content into
> multiple languages, helping to disseminate information globally.
>
> Frequently Asked Questions (FAQ) generation: AI can identify common questions
> related to a topic and generate clear, concise answers.
>
> Metadata generation: AI can automatically generate metadata for technical
> documents, such as keywords, tags, and descriptions, improving searchability
> and discoverability.
>
> Plagiarism detection: AI can identify potential plagiarism cases in technical
> writing and suggest alternative content to maintain originality.
>
> However, it is important to note that ChatGPT and similar AI models are not
> perfect and may still require human supervision, especially in highly
> specialized or complex fields. Human writers will continue to play a critical
> role in the technical writing process, ensuring accuracy, relevance, and
> context in the content produced by AI.
## Potential productivity boon

Long story short, I can imagine a future where generative AI does not replace
technical writers but is instead just a powerful new tool that increases our
productivity. There are lots of possibilities in this regard, but I will focus
on only one for now.

Take the age-old "never enough time to document everything" problem for example.
Once I become proficient in the technical domain I'm documenting, I often
realize how little of that domain I will ever be able to document. There simply
is never enough time to cover everything. Generative AI has the potentially to
dramatically speed up the research and content creation process, so it has the
potential to help me finally make progress on the "never enough time" problem.
Generative AI doesn't really seem like a threat to me in this context for a few
reasons:

* It's work that I never previously had time to get to anyways.
* It may still be more efficient to have a technical writer in the mix than
to try to automate everything completely. After all, as xkcd so brilliantly
summarized in [Automation], the funny thing about automation is that you do
it because you think it will save you time when in reality it's often the
opposite.
* You may still need someone who is proficient in technical writing and
the technical domain in order to effectively orchestrate the generative AI.

## The jury is still out

I caution against taking either side at this point. It's simply too soon to
know yet.

## Summary of this post from ChatGPT

Here is how ChatGPT summarized this post:

ChatGPT (GPT-4):

> Generative AI, like ChatGPT, is seen as both a potential threat and a productivity
> boon to technical writers. While it could reduce the demand for writers as AI becomes
> increasingly proficient at their tasks, it could also be a valuable tool to enhance
> productivity by speeding up research and content creation. In this context, AI could
> help address the "never enough time" problem in documentation. However, technical
> writers might still be needed to effectively orchestrate generative AI. It is too early
> to determine the exact impact of generative AI on the field of technical writing.
Here was the prompt I provided:

> Please summarize the following text in 100 words or less:
>
> `<content>`
`<content>` was a copy-paste of the published web content. I omitted the
"ChatGPT's take" because the ChatGPT UI has a word limit.

[generative AI]: https://en.wikipedia.org/wiki/Generative_artificial_intelligence
[ChatGPT]: https://en.wikipedia.org/wiki/ChatGPT
[general purpose technology that may affect most jobs]: https://twitter.com/emollick/status/1637790373689917441
[code]: https://twitter.com/emollick/status/1637961568238411776
[image]: https://twitter.com/emollick/status/1639094707795165184
[Lighthouse]: https://developer.chrome.com/en/docs/lighthouse
[Automation]: https://m.xkcd.com/1319/
<script>
window.location.replace('https://technicalwriting.dev/genai-outlook-2023.html');
</script>

0 comments on commit 48be26a

Please sign in to comment.