AI is on everyone’s lips. In particular, OpenAI’s ChatGPT service, which was released last November, is currently enjoying great and increasing popularity. It is now undisputed that AI will change the nature of work, especially in the communications industry. But will we have all our texts written by AI in the future – or will communication that fulfills its purpose still need humans? We have tested ChatGPT and considered its likely impact. Based on this, provide five reasons why AI will not take the job away from human communications experts . At least not in the foreseeable future.
1. No creation of new content
AI is an amazing new technology: it can compile texts, provide answers to questions or give tips on everyday challenges. For this purpose, the tools use the Internet in particular and collect their information there. Although this results in newly compiled texts from various sources, the content is not really new. The AI only knows what you have previously told it and what you have fed it or what it is feeding itself from the internet. This may be sufficient for short, simple texts or tweets, but not for more extensive texts such as media releases or even communication concepts. As part of personal contact with our customers, we get to know their needs and can respond to them individually. AI, on the other hand, treats all texts equally, so there is no personal touch. However, when it comes to reaching the target groups as precisely as possible, this point is relevant – because it ultimately helps a message to be successful. So, you can’t expect to get “cut through”, or differentiate your service or products with such text.
2. Sources need to be verified
Services like ChatGPT collect their information randomly from the internet. Cases have already been reported in the past in which ChatGPT even made racist remarks. ChatGPT developer OpenAI has already reacted to this and, according to its own statement, solved the problem, but it remains unclear what the software uses as sources. Anyone who uses ChatGPT must be aware that not all information that is spat out is correct. References are often not available, and when they are, they are frequently incorrect. And looking up sources afterwards is practically impossible. Since references are essential for researching and verifying information, the use and benefit of ChatGPT texts is relatively limited.
3. Who owns the copyrights?
The question of who the author of a text is created by an AI needs to be considered as it presents a risk. According to OpenAI, the content belongs to the company itself. Since ChatGPT uses a wide variety of sources on the Internet, copyright infringements by users is practically permanently possible.
4. Working with ChatGPT does not actually save time – if you want quality
The groundwork provided by tools like ChatGPT can no doubt be helpful. If, however, a considerable amount of time must be invested in developing clear messaging, a unique proposition as well developing and editing texts that truly differentiates your offerings to key stakeholders Based on experience with clients who have clearly targeted audiences and specific offerings to them, there is no time saving with the use of AI.
5. It can’t replace personal contact
Personal contact (the “human factor”) is an important, often even decisive, component in effective communications. It allows us to express emotions, ask questions and obtain answers that go much deeper than than an AI enabled text. Direct contact also enables us to identify and correct issues and misunderstandings before they escalate. While no text can replace personal contact, the promise of AI has the potential to be seen as an easy solution to corporate costs pressures, especially in sales and marketing. Tools like ChatGPT cannot replace human interactions. Subtle cues like gestures, intonation, body language, and reading between the lines are imperceptible to the AI and cannot be incorporated into the text.
OpenAI is working to further improve their AI offering. The company recently introduced GPT-4. Compared to its predecessor, this version should be able to answer even more creatively and precisely, understand longer, more complicated information sources and process images as well as texts. The current version of the AI is already good enough for smaller tasks such as writing a simple social media post, but the current level is not yet sufficient for longer, subject-specific texts. But we are curious to see what else will be possible with AI technology in the near future. For this reason, we are continually testing it to find the optimal resourcing solutions for our clients.