In a world where AI tools such as ChatGPT, SwissGPT or Peak Privacy are becoming more and more powerful, not only are our tools changing, but also the way we work. AI assistants such as ChatGPT, Claude or Perplexity can be a game changer, especially for creative or strategic tasks that start on a “white sheet”. But in order to exploit the full potential, we need to adapt our working methods and process steps. This article shows what a new AI-supported process could look like and what opportunities, risks and requirements are associated with it.
Why Prompted AI Collaboration?
The term Prompted AI Collaboration combines two central aspects of the new work process: Firstly, the crucial role of the prompt – i.e. the detailed task description that serves as the basis for the AI. Secondly, the collaboration between humans and machines or between humans and artificial intelligence, in which the strengths of both partners are optimally utilized. This name emphasizes that the success of this process depends on the clarity and precision of communication. The inclusion of “AI” in the title also makes it clear that this is an innovative approach geared towards the use of AI tools.
The new process: AI-supported work organization
1. clarify the initial situation and objectives
Before an AI assistant such as ChatGPT, SwissGPT or Peak Privacy can be used sensibly, it is essential to be clear about the initial situation and the objectives of the task. What is to be achieved? What are the general conditions? This clarity forms the basis for all further steps.
2. create a comprehensive prompt
Successful use of ChatGPT, Peak Privacy and the like stands and falls with the quality of the prompt. A detailed prompt describes the task, the objectives, the context and any style guidelines. It is helpful to provide concrete examples or structures.
3. check and refine the result
The first result from ChatGPT, Claude, or similar is often a rough draft. This is where the dialog phase with the AI begins: providing feedback and making adjustments to iteratively improve the results. Incidentally, the Canva function of ChatGPT can be very useful here. I explain the Canva function in one of my last videos.
4. final self-examination
Once the result has been optimized, a final critical check is necessary. Is it in line with the original goals and expectations?
5. obtain team feedback
For team tasks, the results should now be shared with colleagues. Your feedback helps to identify blind spots and bring in new perspectives.
6. make adjustments
If there are many suggested changes, it may be useful to revise them again in collaboration with ChatGPT or SwissGPT or other AI assistants. If the adjustments are minor, they can be implemented directly.
7. final examination by ChatGPT and Co.
The final result can be evaluated again by ChatGPT or a similar assistant. This involves checking how well it fulfills the original prompt and the objectives. It is also possible to change the assistant or the language model (LLM) again here to change the view.
Practical example: Development of a new course program for nursing staff
1. initial situation and objectives:
A training provider wants to develop a new course program for nursing staff. The aim is to offer training courses that meet the current challenges in nursing care.
2. create prompt:
A possible prompt could look like this:
“Imagine you are an education expert. Develop a course program for nursing staff. The courses should cover current topics such as digitalization, patient communication and stress management. Please suggest five courses, including description, target group and learning outcomes.”
3. check the result:
The course program proposed by ChatGPT is analyzed. Is the content relevant? Do they sufficiently cover the target group? Are there any areas that are missing?
4. final self-examination:
The program is critically scrutinized and checked for completeness and feasibility.
5. team feedback:
The program is discussed with colleagues from the care and education sectors. Their practical experience helps to further improve the program.
6. adjustments:
If appropriate, the changes will be implemented again by ChatGPT, Peak Privacy or SwissGPT.
7th final examination:
The final course program is evaluated by ChatGPT or another AI assistant to determine how well it meets the original requirements.
Comparison: classic vs. AI-supported process
| Classic process | AI-supported process |
|---|---|
| Idea generation and concept development are purely manual. | AI supports the creation of initial drafts. |
| High time expenditure for research. | Faster results thanks to comprehensive AI databases. |
| Teamwork is often time-consuming. | More efficient teamwork thanks to pre-structured results. |
| Greater susceptibility to errors in structuring. | AI provides clear structures and suggestions. |
Opportunities and risks
Opportunities:
- Increased efficiency: AI-supported processes can significantly speed up time-consuming tasks. Employees save valuable working time and can concentrate on strategic tasks.
- Quality improvement: AI helps to deliver structured and well thought-out results. It identifies potential gaps or weaknesses that would otherwise be overlooked.
- Creativity: AI can introduce unexpected perspectives or innovative solutions that expand the thought process.
- Scalability: Projects can be scaled with less effort, as AI processes large amounts of data efficiently.
- Flexibility: The adaptability of prompts allows the AI to respond to different tasks and requirements.
Dangers:
- Dependence on AI: There is a risk that employees will rely too heavily on AI and neglect their own critical thinking or problem-solving skills.
- Data protection: Entering sensitive information in AI tools harbors risks, especially if the data is not adequately protected. Data protection-secure tools such as Peak Privacy or SwissGPT should always be used when using sensitive data.
- Lack of originality: Results generated exclusively by AI can be perceived as generic or unoriginal.
- Misinterpretations: If prompts are formulated imprecisely, the AI can deliver results that do not meet expectations.
- Complexity of implementation: Integrating AI into existing work processes can be time-consuming and often requires additional resources.
Prerequisites:
- Training: Employees need to be trained to use AI effectively. They should understand how to optimize prompts and critically evaluate the results.
- Clear processes: The workflow should be clearly structured in order to optimally integrate the use of AI. It should be clearly defined when and how AI is used.
- Technical infrastructure: Secure and fast access to powerful AI tools is essential. A robust IT security architecture is just as important.
- Cultural acceptance: An environment must be created that is open to new technologies. Skepticism towards AI should be reduced through education and transparent communication.
- Regulations and guidelines: Companies should develop guidelines for the responsible use of AI in order to comply with ethical and legal standards.
Conclusion
The integration of AI assistants such as ChatGPT marks an organization of work in which efficiency, quality and creativity can be redefined. Tasks that need to be developed from scratch in particular benefit from AI’s ability to quickly deliver structured and well thought-out designs. This not only saves time, but also opens up new opportunities to pursue innovative and unconventional approaches.
But the use of AI is not a sure-fire success. In order to maximize the opportunities and minimize the risks, several prerequisites are crucial. Employees need to be trained, not only in the use of AI tools, but also in the ability to critically scrutinize and refine results. Establishing a clear process structure that defines when and how AI is used is just as important as access to a secure and efficient technical infrastructure. At the same time, companies need to promote a culture that is open to new technologies and establish ethical guidelines for dealing with AI.
The dangers, such as dependence on AI or the potential neglect of human creativity and critical thinking, should not be underestimated. This makes it all the more important to view AI as a complementary tool that supports, but does not replace, human skills. Employees who master this balance will be able to make the most of the benefits of AI and develop their own skills at the same time.
The future belongs to those who are able to combine the strengths of man and machine. Prompted AI Collaboration is an example of this symbiosis, in which precise human input is enhanced by the performance of AI. This approach can not only make processes more efficient, but also open up new opportunities for collaboration and problem-solving. However, this requires a conscious and responsible approach – both in the implementation of the technology and in the development of the necessary employee skills. This is the only way to exploit the full potential of AI and achieve sustainable results in the long term.
The process applied
This blog post is an example of the useful application of the new Prompted AI Collaboration work process and was created using exactly this method.
The output prompt in detail
I want to write a blog post that shows the new way of working with AI assistants like ChatGPT.
My thesis is that for many tasks, we need to change the work organization or the process steps.
This applies above all to tasks where we start on a blank sheet of paper, so to speak, i.e. where little exists so far. For example, a rough concept for a new strategy, a blog post or similar.
My new approach here is as follows:
- Becoming aware of your own starting position and goals.
- Write a comprehensive prompt for ChatGPT that describes the task in great detail and asks ChatGPT to complete this task
- Check the result of ChatGPT very carefully and, if possible, enter into dialog with ChatGPT and make improvements. The Canva function of ChatGPT is suitable for this.
- Question the final result of ChatGPT again yourself.
- If it is a team task, now is the time to involve your team colleagues and get their feedback.
- If there are a lot of customization suggestions from human colleagues, it may be useful to make these adjustments again in cooperation with ChatGPT and to repeat the process described here. If the adjustments are rather minor, they can also be added to the previous result without ChatGPT.
- The final result of ChatGPT and humans could then even be evaluated again by ChatGPT at the end and checked to see how it matches the original prompt and objectives.
I would now like to write a blog post in which I describe and explain exactly this new way of working.
There is first an introduction that introduces the topic.
This is followed by a description of the new process.
The new process is explained directly using an example. One example is the development of a new course program for nursing staff.
The classic process is then compared with the new AI-supported process.
This is followed by a section on opportunities and risks. The advantages of the new process are shown, but also the risks and the necessary requirements.
The conclusion provides a summary and also looks at how employees’ skills will have to change in the future.
You are now writing me this detailed and very long article and also developing the examples mentioned. Of course you write the post SEO optimized.
At the end, come up with a name for the new AI-supported process and explain how you came up with it.
Take enough time to ensure that the result is really good.
I think the chosen title is good, but somehow it needs to be shorter and more concise so that it quickly sticks in the minds of the target group. and it would be good if it is an English formulation.
Any further questions?
Do you have any questions? Or would you like to integrate this method into your team as well?
I will be happy to support you, act as a sparring partner and answer your questions. I am always happy to receive your messages, preferably by WhatsApp message or e-mail.
This article is also available as a podcast episode
Attention! The podcast was created entirely by my AI-Assistant based on my contribution – no guarantee for incorrect content.
Need more information?
If you would like more information about the Prompted AI Collaboration process, please fill out the form below.
