Hello from Hamburg,
In the June issue of his newsletter, Holger Volland addressed—under the title “Fear Not. It’s Just New”—the anxiety that typically grips us when technological progress brings about something new and, as with ChatGPT, makes it accessible to everyone. One of the fears centers on the question of what AI will do to our jobs. Has the end of work arrived?
The first credible studies on this topic focus on one of AI’s capabilities: making predictions based on a dataset—that is, translating information we have into information we don’t have. The Goldman Sachs study, for example, concludes that by 2030, up to two-thirds of all job roles in Europe and the U.S. will be affected by this form of AI. The researchers further estimate that in the U.S., one-quarter of all jobs could be automated by AI, particularly white-collar jobs. This would affect standardized tasks such as accounting, basic legal advice, administration, and those professions that primarily involve gathering, reorganizing, and structuring information. The study by the University of Pennsylvania, in collaboration with researchers from OpenAI, even estimates that for 80 percent of all American knowledge workers, at least one in ten tasks could be replaced by AI.
In his blog post in March 2023, Bill Gates envisioned an AI-powered personal assistant that reviews emails and sorts them by importance according to pre-set parameters, knows all upcoming meetings, reads the necessary information and documents in advance, attends the meetings, writes the minutes afterward, and generates to-do lists. What an enticing idea! Personalized marketing will also soon become smarter than it is today, when an e-commerce platform tries to tempt me into buying a washing machine even though I purchased one just four weeks earlier. Doctors will be able to rely on a medical history generated by AI from a digital health record, critically evaluate the machine’s automated diagnostic recommendations, and focus entirely on caring for the patient. Architects will enter relevant information about a building’s parameters, the desired materials, and key design elements into an AI-powered tool and receive initial 3-D design drafts, which they will critically analyze and refine. Lawyers will no longer draft contracts but will instead focus on advising clients during tricky disputes with the opposing party.
There is a pattern that has held true for all technological leaps in the past, as a labor market researcher at the Institute for Economic Policy at the University of Cologne confirmed to me (as, incidentally, did the rule that people are always afraid of losing their jobs during technological leaps):
First: Support roles are being replaced. Just like the profession known as “computer.” In the 1940s, this role was primarily filled by female assistants who performed repetitive calculations in fields such as astronomy or ballistics. These tasks—along with the job title—were later taken over by the “computer” as we know it today.
Second: Many people whose jobs are initially replaced by automation will find a place and a sense of fulfillment in new professions created by AI. Take, for example, the journalist who, as a “prompt engineer,” helps clients ask the AI smart questions in order to get useful answers. Virtually all technical innovations have given rise to new professions and generated lasting gains in productivity. This was true of the breakthrough of the steam engine in Prussia’s coal-mining regions in the mid-19th century, just as it was with the introduction of the personal computer. Initial estimates predict that AI will increase global annual gross domestic product by seven percent.
Third: Most of the affected professions will be supported by AI rather than replaced by it. And this will remain the case for the foreseeable future. Because there is one thing AI won’t be able to do for a very long time, as I learned a month ago from the top AI professors at Stanford Business School in Palo Alto: It cannot empathize with people. This means that critical and creative thinking, emotional intelligence, and active listening—which enhance one’s own judgment—will remain irreplaceable human abilities compared to machines for a long time to come.
Fears about the impact of AI on the job market are unfounded. Instead, we should focus on whether the next generation is being prepared for the changing job market. Mark Siemons of the *Frankfurter Allgemeine Zeitung* recently compared the recommendations of the Standing Scientific Commission of the Conference of Ministers of Education regarding the core curriculum for German schools with the capabilities of language processing systems. The sobering conclusion: Because the standardized and measurable competencies that the current education system is geared toward will likely soon be performed better and more comprehensively by computers, the current curriculum may be preparing the next generation for unemployment.
Back to Holger Volland's newsletter: He expressed his delight that, when it comes to ChatGPT, his readers are not fearful but rather critical and enterprising. That's great.
