Recently, there has been plenty of rather sensational news about artificial intelligence (AI) and how it will change work. Some have even argued that up to 300 million full-time jobs will be affected. But that does not mean that 300 million jobs will simply disappear.

Given the history of capitalism – with for examples, the replacement of steam by coal, the transition from horse carts to Ford’s motor car, the move from diesel engines to electric motors, the change from human’s manual calculations to desk calculators, and eventually the introduction of computers and AI in the workplace – virtually none has led to the end of capitalism and global mass unemployment. 

Yet, AI still has the potential – and many are convinced that it actually “will” – lead to a rather fundamental change in work and, subsequently, influence millions of jobs. In other words, AI will not so much as “kill” the jobs, but it will “change” the jobs. Currently, two key questions are hotly debated in Germany:

1. what opportunities does the use of AI offer?
2. how will AI change the world of work?

Also in Germany, the most controversial AI tool remains ChatGPT. On this, many German IT professionals and companies expect a real upheaval. Yet Germans think that the effects of the recent technology earthquake (as it is often called) are felt globally and no longer just noticeable in Silicon Valley and related industries.   

In addition, most German experts are convinced that the working life of many people has already been fundamentally changed because of the introduction of computing – not to mention future changes that will come with AI. To investigate this, Germany undertook two key studies that illuminate the consequences of the so-called “AI revolution” in workplaces.

The first study was conducted by the developer of ChatGPT themselves. On that, Germany’s news agency DPA reported that ChatGPT’s mother company OpenAI (read: not so “open”) has teamed up with the University of Pennsylvania to find out which jobs are most influenced by ChatGPT.

The study has found that the occupational groups that are particularly affected are those of the: interpreters and translators, survey researchers, poets, lyricists and creative writers, animal scientists, public relations specialists, writers and authors, mathematicians, tax preparers, financial quantitative analysts, and web and digital interface designers.

Germans also believe that accountants are among the professional groups most affected by the possibilities of AI. Partly because ChatGPT could do at least half of many highly standardized work tasks in accounting, and it can do it much faster and with higher accuracy.

Yet, such AI systems can also generate incorrect facts in their answers. As a consequence, there will still be a need for humans to do some fact checking. For example, on “Thomas Klikauer” ChatGPT generated four errors: firstly, I am not a sociologist or, secondly, a professor; thirdly, never co-authored a book “The Management Theory of Frank Stilwell”; and never written a book called “Megacities: Our Global Urban Future.  

Beyond all the media hype, AI often generates a  hallucination-like erroneous facts in their answers. And ChatGPT delivers these wrong facts with great confidence. It can produce completely wrong results in translation, classification, creative writing, and the generation of computer codes. ChatGPT can just make things up – including faked citations.

Despite such obvious errors, mistakes, and made up inaccuracies, algorithm-guided AI systems have already achieved, so it has claimed, considerable results particularly in standardized work tasks such as translation, classification, writing, and generating computer codes.

As it stands today, those who ignore the possibilities of AI and do not use it to increase their own performance can – and indeed often “will” – lose their jobs. On the upswing, German experts believe that by using ChatGPT, German workers can gain a lot of time while devoting themselves to other more creative and more supervisory tasks.

However, there are also professions in which artificial intelligence plays only a minor role – for example, for cooks, car mechanics, and jobs in oil and gas production, as well as in forestry and agriculture. In other words, AI will not cut your hair, be a waiter, and change the wheels on your car.

In a second study which was undertaken by the investment bank Goldman Sachs, they have found that there will be effects of what is called ”generative AI” on the labor market. Many see generative AI systems (GenAI) as algorithm-based computer programs that are able to create new ideas, new content, and even new solutions instead of just working through predefined rules and standardized instructions.

Goldman Sachs rather sensationally also claims, GenAI “could” lead to significant disruptions in labor markets. They argued that around two-thirds of all current workplaces are already exposed to some degree of AI automation and generative AI. This could replace up to a quarter of current work tasks.

Meanwhile in Germany, Hinrich Schütze (Uni-München), has described the development of GenAI as a revolution. He thinks that GenAI is, in terms of technology, comparable to the invention of the Internet and smartphones.

Nevertheless, the consequences are already enormous, there will be big changes in how we write, whenever we write texts, how we program, says Schütze. This also has major consequences for everyday work.

At the same time, however, the AI expert warns that artificial intelligence shouldn’t be given too much scope for action when making decisions. For example, in the judiciary, medicine, tax consulting, and asset management, AI can have dangerous consequences. A key problem is that many people tend to think that it must be true if the AI model is so.

Yet in reality, AI models cannot – yet! – assess their own safety. This remains a very serious problem and much of this directly relates to data protection. Many German AI experts suggest that caution should also be exercised.

Simultaneously, the use of AI remains also important for Germany’s small and medium-sized enterprises (SME). A recent IFAA trends’ assessment showed that the “inhibition threshold” for using AI in SMEs is decreasing.

In their survey, Germany’s IFAA or “arbeitswissenschaft.net” found that 31% of executives believe that AI will become more important in 2023 and beyond. Yet, they also believe that SMEs need to prepare themselves for the integration of AI applications into their operations.   

According to IFAA, employees also believe that AI will continue to gain importance in the future. Yet, they also fear an increase in work-related psychological stress. 

In IFAA’s opinion, this shows that focusing on technical concepts alone is not enough. Therefore, transparency, participation of those affected, and the use of learning concepts are important. Meanwhile, many people aren’t prepared for the influx of AI. In fact, they remain cautious about the use of the new technology. 

This applies to the area of robotics as well. At a German company called KUKA, robots will play an ever-increasing role in the working life of many in the very near future. This also means that today’s robotics’ specialists will have to be prepared for AI-guided robots. 

Similarly, Germans are certain that AI will be imperative in these two areas: the operating environment and in business programming. Quite apart from KUKA, many technical specialists in Germany will have to familiarize themselves with the topic of robotics guided by AI. Germany’s IFAA also thinks that AI technologies can best be described by outlining six components:

• methods and procedures that enable technical systems to perceive their environment; to process what is perceived; to solve problems independently; to make decisions; to act; and – and this is the key –  to learn from the consequences of these decisions and actions.

In any case, it remains imperative to understand that AI does not – yet! – developed human abilities. AI simply performs very specific work tasks. For example, AI is particularly good at processing large amounts of data. It works even better when these data demand a very high computing capacity that no human being could process during a given timeframe. In essence, AI is a digital tool.

During the past months, various potential applications for AI have been reported and discussed extensively in German media – almost at a daily level. For example, AI is forecasted to have many applications in Germany’s medical and public health area.

Despite this, IFAA is convinced that AI will become the defining universal technology of the century. It argues that AI offers plenty of opportunities for innovative models of companies and state institutions. As a consequence, work is likely to undergo radical changes, with at least, four key issues: 

1. the introduction of AI instruments;
2. AI-assisted work systems;
3. learning robots; and, 
4. the provision of user-optimized information.

For employees currently working in German companies operating any one of these four areas, AI will mean more flexibility and more demands on work tasks. Simultaneously, removing monotonous and routine activities through AI will facilitate this. Three manufacturing examples exhibit this already:

• One of the major German manufacturers of sensor technology – Balluff GmbH in Neuhausen – has been able to make very complex analytical problems manageable through the use of AI;
• Drehtechnik Jakusch GmbH in Saalfeld – a manufacturer of metal and plastics – used AI to develop a multi-faceted and innovative shift planning system. This made it possible to plan idle and standby times more sensibly.
• At SICK AG in Waldkirch, new AI technologies such as deep learning are being used. With this, the company has developed innovative machine vision systems for applications in logistics.

Of course, there are many more examples in Germany on how AI can be successfully used in practice and how it speeds up and facilitates work. In corporate and business practice, German manufacturing focuses on the following AI applications: predictive analytics; optimized resource management; quality control; knowledge management; intelligent assistance systems; robotics and autonomous driving; intelligent automation; and intelligent sensors.

Recent evidence from Germany clearly shows that many companies have at least some “theoretical” knowledge about the use of AI. Yet at the same time, they lack practical applications. It is for this reason that AI examples are important in order to show how AI can be developed in practice and how German companies can use their resources to develop new forms of work organizations based on AI.

On the downside, it seems there is still a long way to go in Germany before AI is used in workplaces. One thing that is perceived as a certainty, is that this path can no longer be avoided. 

For that, each company will have to decide for itself on how AI will change work and this means that managers, trade unions, and employees must work together. Meanwhile, the current media hype about AI has raised hopes for what Germans call “a new golden information age”. 

However, many German manufacturing workers also fear that AI can be harmful and that AI will at least have a major impact on their workplace. And this only in the wake of  ChatGPT and AI-guided new software. 

Despite these problems, German workers and company bosses are convinced that artificial intelligence will change the way they work. All three – trade unions, employers, and the state – seem to agree that there are also potential dangers and that Germany – and indeed the EU – needs to regulate AI. 

On the bright side, Germany is already planning a KI-Verordnung – an administrative AI regulation below the level of a law. Simultaneously, Germany is also supporting an EU regulation for AI at European workplaces. 

At both, the German and the EU level, Germany’s powerful trade union peak body – the DGB – supports AI regulation within Europe’s participatory working structure of labor relations. On inclusive labor relations, Europe still leads the way.

---