כותבים טכניים

Technical writing, in many cases, is already extremely structured and stylistically boring. A good technical writer writes like a robot. Consistency, controlled language, recurring text snippets, pre-defined text structures, etc., all make automation very easy.
The main challenge for an AI is not the technical writing. It is actually making sense of the source material. This will be the main job of technical writers: Explaining to the AI what it needs to write. Technical writers will increasingly be busy with post-editing AI written texts, which will be a very mind numbing task. It will also require fewer people and warrant lower pay.
If you want to know the future of technical writing, look at what has happened to technical translation. If you are a technical writer, you are headed the same way. You will operate AI tools and check the results for lousy pay because you are not actually doing the brain work. Get into nursing or open a funeral home. People will always get sick and die, that's a future-proof occupation.

Technical writing only works if the resulting texts fit the target group.

As technical writers we understand complex systems (as well as the engineer's gibberish that often comes with it) and are able to translate this into a form that is easily understood by the respective user groups, i.e. operators or maintenance personnel.

In my opinion AI nowadays can provide draft text to the technical writers. But it takes a human to process that into understandable content, because only a human can know what is necessary for others to profit from precise instructions and related warnings. Also when it comes to jurisdiction.

Who will be held accountable for injury or death caused by automated AI that simply pretends to be human-like but has otherwise no conscience? I mean, an AI can tell you that a stove is hot, but does this mean it really knows the consequences of touching it?

Some learned it the hard way, I doubt an AI can.

When I came to this profession I was actually surprised how little writing it involved. Even if AI replaced the writing bit of my work completely (say, it would be able to create perfect content from whatever input I feed it - at which point most white collar jobs are doomed anyway, btw), I'd still have like 80% of my current workload to deal with myself. Coining job titles is elusive nowadays, but I'd say think bigger: technical communicator, information architect, information designer, content strategist, etc. I've done things from under those labels and much more - from simple coding to graphic design and basic video editing. It is kind of a 'jack of all trades' profession, or at least it can be, but in a good way - opens up many doors, and you could easily switch your focus elsewhere once part of your job is automated. Nobody is safe from automation, but here at least you have flexibility and transferable skills. '79%' is a joke. I transitioned to tech writing from a profession rated much lower, for the very reason that said profession is being automated out of the market right before my eyes. So take info on this website with a grain of salt. ;)

I am a technical writer. And I work on AI. We do NOT have a high number of tasks easily automated. Can an AI tool write the first draft? Yes. Can AI do translation jobs? Yes, we already use machine translation. But the bulk of what we do with edits, technical reviews for accuracy, no AI isn't going to take those jobs away. Like all technology, it might reduce head count and there will be fewer jobs. But these jobs are not going to disappear.

This confirms that most people do not know what technical writers do. There is a lot of interdepartmental discussions and planning that cannot be automated.
I don’t see this role being completely automated in 10 years.

Technical writers do not do just menial, repetitive, tasks that do not need any judgement. The job requires extensive knowledge and judgement something that AI cannot do.

Explaining, proofreading, summarizing, are all jobs that even ChatGPT4.0 can do well. Once LLMs are available in local versions that don't reveal confidential info to the public, and have advanced a bit, technical writing will be generated mostly on-the-fly from user questions.

Curating and organizing the information to feed to the AI, and checking their output for accuracy, may be one of the few jobs left to tech writers.

Who will write about AI systems? AI itself?

While the promise of AGI is still remote, and LLM AIs are not great at giving accurate responses, we cannot rule out that there is a toolset which could do this job and one that somebody can come up with within the next 20 years. 20 years is a lonb time to refine a toolset.

Because management/shareholders care more about dollars than quality, and apparently customers are getting used to it. Technical writers constantly fail to prove their value (with data), and are thus seen as not adding value. Technical documentation is seen as a 'gimme' (expectation of free resource that comes with the product), and thus valueless or worse.

If AI can perform well enough and at a lower cost, our expertise will lose its value. In many offices, we are already considered a luxury. It will be even harder to justify our worth if AI can make the writing process less time-consuming and painful for engineers. Many jobs, not just technical writing, are seen as luxuries by employers who lack the necessary skills or time to do them properly. If AI can reduce time and costs while producing reasonable content with minimal input, what will be left for technical writers to do?

More importantly, why would anyone pay for a professional to do it?

It's like saying teachers won't be needed because we will have textbooks to study from. Except we are talking here about replacing authors of textbooks (not to mention the fact that textbooks already exist and teachers are still there).

More and more technical documentation has become structural writing. Writing documentation is more and more like writing a code. AI will learn writing a code easily in the future.

There’s a massive human element to tech writing, but never underestimate technology and where it might be in five years time. Also don’t ignore wishful thinking and subject ignorance.

Technical writing goes well beyond just writing the actual sentences. I cannot fathom how AI would somehow be able to do all the things that are required to be done in order to complete a technical writing piece. AI has already taken over the writing niche . . . people no longer need education to write, software helps them do it. That is all AI could ever do for a writer.

Until both the end-user and the engineer are both automated there is no feasible way to fully automate this job. You can have some IT tech writer positions that become more efficient through perhaps some auto-text generation but expecting this to translate into any hardware based product is extremely difficult to automate

We already have artificial intelligence which is capable of blogging and reporting news articles without human beings being capable of detecting that this information has been written by what is effectively an algorithm, so it is incomprehensible (to me) that within the coming decades, for the concept of what it is to be a “writer” to remain unphased. I myself have worked as a writer and author and I can emphatically state that artificial intelligence is going to usher in a paradigm shift for those who are currently connected to the field of written work. The profession will remain intact until the end of the century, but swift and vast changes are to be had; this is an inevitability which we must accept, embrace, and use to empower modern day writers

Every company has their own standards for writing. The more successful writers are able to inject some personality into their writing.

How will they take the job if you need to know the systems that you are writing about?

Writing requires creativity and contextual understanding of a particular work. Also, the audience of the written work is humans and it requires a certain sense of being able to understand another human to produce work that the humans can understand. So, I'm not really sure if robots would be able to do that unless they reach the level of intelligence that humans have.