Ralf Müller, an employee of DB Systel, has more than 25 years of experience in the software industry behind him. Originally working as a software developer, he has developed his role into a software architect over time. His focus is particularly on the clear presentation and recording of his theoretical approaches. For this reason, he has created docToolchain, an open source project dedicated to the efficient exposition of software architectures. He is currently working on the application of Artificial Intelligence (AI) in software design and will give a presentation on this topic at the iSAQB Software Architecture Gathering. He has experience in using ChatGPT as a tool for solving specific sub-problems in the iSAQB Advanced Level sample exam and as a knowledge repository in software architecture tasks.
Müller is fascinated by the use of ChatGPT to tackle the iSAQB Advanced Level sample exam. As he explains, the entire exam questions were not forwarded to ChatGPT, as the AI tool can only process a limited context. Instead, he used ChatGPT as a training partner to help him find solutions to specific sub-problems and review architectural decisions. It is enough to get a starting point with the help of ChatGPT. By breaking the review into smaller tasks, he was able to create solid queries that led to insightful and targeted responses from ChatGPT.
From November 27-30, the iSAQB Software Architecture Gathering 2023 will be held, which is considered the leading online conference in the field of software architecture. iSAQB and iX, the trade magazine for IT professionals, are responsible for organizing the conference. The program includes a total of 30 presentations in English, including topics such as architecture intelligence for architects and the new accidents of architecture. Ralf Müller will give a presentation on the use of AI in software design as part of this event. Two workshop days will take place prior to the two conference days.
Müller’s experience shows that ChatGPT can not only be used for direct solutions, but can also be used as a tool to identify gaps or contradictions in existing data. Although the result is not necessarily a direct answer to an audit question, it does demonstrate a new approach to problem solving.
Conversational interactions with AI can occasionally lead to unexpected and potentially inaccurate answers. It is therefore important to implement security measures and procedures to ensure the accuracy and reliability of the information and suggestions provided by ChatGPT. Müller believes that the information provided by ChatGPT should be treated as advice from a human expert. He uses his knowledge and skills to quickly check whether what is being said makes sense. If there is a greater need for certainty, he consults official documents or other sources considered reliable to verify the facts.
The question of how large-scale language models will influence our working lives in the future is a recurring one. Ralf Müller does not believe that artificial intelligence will make people superfluous in the future. He describes this idea of technology replacing humans as more of a myth than a reality, as it has already appeared more frequently in the IT world in the past.
The interview on these topics was conducted by Dehla Sokenou and Lukas Zühl.
