By Nils Klute, IT Editor and Project Manager Communication Cloud Services at EuroCloud Deutschland
Anyone working with artificial intelligence (AI) must be aware of its capabilities, properties and limitations. Why do machine decisions need human experience? And how Service-Meister is now training technicians from SMEs on AI.
Computers were supposed to fail, nuclear power plants to go out of control and even civil wars to break out – the Y2K problem put IT to the test. In the software of the time, year numbers were usually written in two digits. A circumstance that can be traced back to the time when data was punched on holes in cardboard boxes: On the so-called punched cards, storage space was tight and space was to be saved – just like with a four-digit year. The real consequences of the Y2K bug were manageable. However, the way people inscribe their reality into machines is causing problems elsewhere today.
ChatGPT: All-rounder and master of discrimination at the same time
Whether answering questions, writing poems or writing complete dissertations – AI applications like ChatGPT are not only true all-rounders, but also masters of discrimination. What is the reason for this? An AI like ChatGPT draws its knowledge about reality from freely accessible and public texts. Texts in which people have inscribed their individual and biased view about the world. The result is whether it’s reservations about women, immigrants or senior citizens – prejudices like these creep into the answers of the bot, as if imprinted on a punch card.
Daniel Munro from the Canadian Munk School has put it to the test. Just one question was enough for the scientist to demonstrate on Twitter just how biased ChatGPT is in terms of male, Western culture. It’s no different with untruths that AI spouts at face value, which may have been part of the reason why managers and entrepreneurs like Elon Musk recently warned about the risks of the technology.
AI project Service-Meister: Including human experience in smart applications
For sure, anyone who works with AI must be aware of its capabilities, properties and limitations. Not to exclude technology, but – quite the opposite – to include human experience in machine decisions. Service-Meister shows why this is necessary. To take industrial service to the next level of evolution with AI, the project has been integrating human know-how into smart applications since 2020. Digital guides, augmented reality applications, apps or chatbots should make expert knowledge scalable. And this is by no means to replace skilled workers, but rather to enable lesser-skilled technicians to carry out complex repairs. The goal is to create an open service ecosystem that closes knowledge gaps and makes German SMEs successful and competitive with AI tools.
Online training courses provide AI background knowledge for companies
What does it take for human and artificial intelligence to complement each other optimally? Firstly, the necessary understanding of how they can be integrated and applied. To ensure that SMEs are increasingly successful in doing so, they plan their AI entry via the Service-Meister website, for example. And secondly, it requires the essential background knowledge of the technology, its opportunities and limitations in the service business.
Service-Meister has now launched its own training programme precisely for this purpose. In several freely available learning platforms, the project qualifies service employees in AI. For example, an e-learning course on the AI Campus shows the possibilities. The online course can be used at one’s own pace and explains, for example, how smart AI chatbots support technicians with repairs. It is no different with the train-the-trainer programme: AI specialists for the service sector can be trained online to pass on their know-how to SMEs in their own courses. The learning courses are based on an open source curriculum from Service-Meister, which is available to educational institutions free of charge.
AI-Know-how: Being able to take the right measures in time
Gain and apply knowledge – not only because it is technologically necessary, but also economically necessary. According to the AI study by the eco – Association of the Internet industry, if the economy as a whole and SMEs use AI across the board, GDP growth of over 13 per cent can be achieved by 2025 (compared to 2019). This corresponds to a total potential of around 488 billion Euro. Regardless of whether it’s smart algorithms or punched cards – being able to recognise risks and take the right measures in good time pays off in both cases. Although displays in Japanese nuclear power plants or American slot machines failed on the night of the millennium change, overall, the evening remained largely inconsequential.
Image credit: iStock-1286627625
Did you like this article? Then subscribe to our newsletterand receive regular updates on similar topics and the Service-Meister project, and discuss this and similar exciting topics with us in our LinkedIn group.