Artificial intelligence is changing the world and has made great progress in recent years: ChatGPT and other applications can perform tasks in seconds that we probably could not perform so efficiently in such a short time.
But will many jobs be lost in the future because machines can do our work faster?
German economist Jens Sidekum sees no imminent risk of mass unemployment due to the continued spread of artificial intelligence. There will definitely be changes, as artificial intelligence (AI) can be widely used. Certain professional fields are in fact characterized by activities that can easily be replaced by technologies. According to the German economist, these activities include “ordinary administrative activities such as filling in Excel files, but also writing increasingly repetitive standard texts, researching and collecting information”. These are all things that AI could possibly do more efficiently. But that does not mean that people who continue to do this work will be out of work because of it. On the contrary, workers will be freed from tedious tasks in the future and therefore have more time for activities that make people indispensable.
Which professions are at risk?
A team of Swiss researchers led by artificial intelligence expert Dario Floreanos investigated which professions are particularly at risk from artificial intelligence. Machines already have dexterity and physical strength, but they’re also surprisingly aware of problems, as they can tell when something isn’t going as planned. What machines lack is originality, coordination, or problem-solving ability. Armed with this knowledge, the researchers calculated an automatic risk index for each occupation.
The butcher trade is the most threatened. It has an automation risk index of 78%. This means that the robots already have 78% of the necessary skills. At the other end of the spectrum are physicians. The hazard ratio is 43%. Today, machines already have nearly half of the skills that make up perhaps the safest profession.
According to the danger index, jobs such as engineering pilots, air traffic controllers and most medical professions are safe. An exception is radiology specialists. They are already in the middle because artificial intelligence can do some of the diagnostic work. However, this is where the study reveals a weakness: the database lists 18 essential skills for GPs – empathy is not one of them. Researcher Raphael Lalive says they focused on basic physical and mental skills. They would not have captured all the reality of the profession, but at least a very large part of it. Models could be in trouble. In the middle of the lower ladder are the bartenders and orderlies. The jobs of cashiers, dishwashers, taxi drivers or models, now replaceable by digital, are even less secure.
Are journalists in danger?
The OpenAI researchers who developed Chat GPT at the University of Pennsylvania, on the other hand, sometimes come up with different predictions than the Swiss researchers. According to their study, people in these professions should be prepared for the fact that artificial intelligence can take over at least some of their previous tasks: programmers, mathematicians, accountants, interpreters, writers, journalists. Chat GPT already has remarkable results in tasks such as translation, classification, creative writing and computer code generation. Especially in the field of journalism, however, one should not leave the business entirely to ChatGPT as it lacks the ability to judge the facts. The American researchers assume that most jobs will be modified in one way or another by AI language models. About 80% of American workers are in jobs where at least one task can be done faster with AI. But there are also professions in which artificial intelligence will only play a secondary role. These include, for example, cooks, car mechanics and agricultural jobs.
Even lawyers belong to the group of professions threatened with extinction, since some of its activities can in principle be automated. “Does this now mean that all lawyers will be unemployed? No, probably not. But that means the profession could probably change quite a bit in the future,” says economist Sidecum. If lawyers make smart use of new technological capabilities, they can focus more on really creative things and working with clients. The same is true for other occupational groups.
EU: law on artificial intelligence
Considering the different risks of artificial intelligence, the EU is considering an AI law. The law aims to comprehensively regulate the provision and use of artificial intelligence by private and public entities.
Media ethics expert Dr. Jessica Hizen from the University of Tübingen explains that there are already rules for dealing with artificial intelligence in many areas: for example in stock market calculations, in autonomous driving or in medicine and the surgical robots used therein.
On the other hand, the sector of the “platform economy”, in which companies such as Meta or Twitter are active, has so far hardly been regulated. Hizen doubts that companies regulate themselves. Microsoft recently terminated much of the AI ethics department.
He knows no moral boundaries
What measures does EU law provide for artificial intelligence? To ensure responsible use of AI, the EU wants to create binding rules with a “risk-based approach”. If AI systems fall into the dangerous category of ‘unacceptable’, ‘because they breach EU values such as fundamental rights’, they should be banned. Prohibited practices include, but are not limited to, techniques that subtly manipulate people, causing physical or psychological harm.
It is therefore decisive how the law will ultimately be designed.
Claudia Paganini, professor of media ethics at the Munich School of Philosophy, also sees European AI legislation as fundamentally positive. He finds it problematic, however, that exemptions with arguments such as “national security” create loopholes for AI applications that could potentially be used in undemocratic ways. He cites as an example the issue of facial recognition in public spaces.
Expert Jessica Heesen calls for more rules, especially when artificial intelligence and algorithms are used in communication, especially in the area of social media:
Artificial intelligence systems, which can easily create fake images with a deceptive appearance, have shown the already existing dangers of the technology, says Hizen. The problem is that we will no longer know: what is real and what is not?