Survey on the use of artificial intelligence in technical communication: the results
Artificial intelligence can help create, organize, and retrieve content. But what tasks are best suited for AI, and how is it changing the technical communication profession?
In their presentation at the tcworld conference 2024, "All AI or what? Is AI the ultimate tool for technical communication?" (Alles KI oder was? Ist KI die Wunderwaffe für die Technische Kommunikation?) , Prof. Dr. Britta Görs and Ulrike Parson shed light on these questions. The presentation was given in German.
👉 Before the presentation, we invited people to participate in a survey, and 34 people responded. The results, which Britta Görs and Ulrike Parson already presented in their talk, are now also shown here. The survey was conducted in German. We explain the answers in the accompanying text.
Who participated?
Both senior professionals and relative newcomers to our industry participated.
Use of AI applications
The vast majority of participants are already using AI in their daily work. Question: Have you used AI in your daily work?
- Yes: 28
- No: 6
The most popular tasks AI is used for include: authoring, reviewing and optimizing, summarizing, translating, and generating images.
Company policies on AI
More than 60% of participants said that the use of AI is either recommended or tolerated in their company. Nearly 30% of companies have already established clear policies, and 41% are in the process of doing so.
Frequent tasks and manual/repetitive Steps
If you want to use AI in technical communication, it is important to find out which activities can be supported or even completely taken over by AI. Automation only makes sense if these activities are performed frequently and/or by many technical communicators.
That's why we first asked participants which tasks they perform most frequently.
For many respondents, writing and maintaining text (1) is one of the most frequent tasks in technical documentation. Other activities include communication and coordination (2), information research including standards research (3), project management (4) and quality assurance (5) including terminology checking.
The second question in this area focused on which of these tasks involve many manual or repetitive steps. Tasks with manual steps are candidates for automation or the use of AI.
The frontrunner here was gathering change requests for the documentation (17), followed by expert research (14) and content reuse (13).
Expectations of AI
The majority of participants agree that AI can replace monotonous tasks and free up time for creative or strategic tasks. Many see AI as an aid in finding information and extracting knowledge. Overall, technical communicators seem to have a positive and open attitude toward AI.
Support and training
Many companies have provided training on the use of AI (16), but more than half of haven't (18).
Suggestions for using AI
Participants had many ideas about how they would like to use AI in their day-to-day work. They included:
- Check and correct terminology
- Compare different products and their designs and filter out the differences
- Create draft content
- Automated description of how software works
- Assign metadata
- Research information
Conclusion
The survey shows that AI has great potential in technical communication, especially for tasks such as authoring, translation, and knowledge extraction. Many participants see AI as a way to minimize monotonous activities and promote strategic or creative tasks.
Organizations that establish clear policies early on and train employees to use AI can benefit in the long run, while reducing fears and misunderstandings. AI offers the opportunity to rethink the way we work, enabling greater efficiency and creativity.
Have you had similar experiences or do you have a different opinion? We welcome your comments and discussions!
Feedback of participants
Some participants added comments and questions to the survey. Here are some excerpts:
Who maintains the AI, where does the data come from, who trained the AI and with what? Can the data be used internally or externally?
Risk of AI from my perspective: lack of knowledge about its limits and therefore incorrect use; there should be targeted training on how to use AI sensibly and what AI cannot do or where testing is necessary. But I can only check if I understand how the AI works. How can you convince coworkers that AI will completely replace technical communication?
UX writing is a very common task in technical writing and should not be forgotten.
Have questions about AI in technical documentation? Contact us!