Will AI start writing your internal communications?
Given the recent and ongoing hype over generative AI tools like ChatGPT and DALL-E, professional content producers and corporate communicators have been wondering if such tools might take over their roles to some degree.
In fact, many software companies are already promoting how their AI tools (typically a simple ChatGPT integration) will generate sales and marketing content, write email, produce news stories, and more.
Many larger corporations have prohibited the use of these content generation tools for good reason, considering these artificially intelligent tools simply regurgitate, combine and rewrite existing content works. This creates a liability risk of intellectual property infringement as these tools generally ignore copyrights and plagiarize without (or with incorrect) source attribution.
A good argument can be made that these artificially intelligent machines are faster, more adept, and more responsive to pattern recognition, and more able to create content on an individualized basis in order to nudge behaviors and initiate emotional reactions. You can see AI at work by paying close attention to the headlines in your newsfeeds and the ads you’re being served.
People have been surprised by what these generative AI tools are able to produce, both good and bad. Perhaps you have even read some of these lengthy and apparently well researched articles, with curious headlines, punchy leads, and fantastical images – all produced by AI tools? Surprising too are the hallucinations and false references, and completely made-up and non-sensical written conclusions these tools have been documented to produce.
AI tools have already proven effective at optimizing digital content to generate clicks and emotional reactions, with the corresponding evidence (and ongoing training of the AI models) advertising impressions and click-through. Yet these communications serve mostly to distract people, given the objectives of marketing and sales communications are to grab and redirect the audience’s attention.
When it comes to internal communications with employees, most companies will not benefit from keeping employees distracted. Most employees have valuable work to do, roles to play and tasks to complete, and delivering communications optimized to distract will result in a reduction of real employee productivity, despite increasing your clicks and response metrics.
Generative AI requires source material to base the new content on. In terms of corporate content, generative AI has limited utility due to the unique nature of announcements, company news stories, policies, executive strategies, process changes, and other communications. That said, digesting a set of corporate communications can establish a “voice” or style of writing and speaking with specific nomenclature and word-choices. And once an AI “learned” this voice, it could provide editing services to reproduce it within any new piece of content.
How will communicators use AI in the near future?
Here are five ways you can expect large language models to help internal communicators.
Summarizing
Plenty of corporate content and industry information is long and complex. Rather than spending hours reading it, ask an AI to provide a one-page summary with a bullet list of the most important points, and you can then dive into anything you want more detail on.
This can also work in reverse. When you have compiled a set of long form content for employees, why not have the AI produce a summary with bullets to help you get the key messages across with links into the details for those inclined to learn more?
Research Compilation
Adding AI to search engine capabilities means you can prompt an AI to not only research specific topics, but to read and summarize it for you. Prompt AI to research topics, answer questions, or gather how-to videos.
Writing and Editing
Beyond grammar and spell checks, an AI may help you write better emails, intranet posts, stories, and other assignments. Ask an AI to reduce three pages to one, or to improve your Flesch reading ease score.
Language Translation
Need to produce or provide content in multiple languages? LLM AIs are able to provide word, sentence and entire document translations for you.
Visual Content Production
Beyond text, AI tools can produce images and video trailers from existing content.
