The staggering development of generative AIs like ChatGPT promises to be a real revolution for the world of work. In particular, it risks making certain professions obsolete, particularly in the tertiary sector.
Since its release, ChatGPT has been eyeing the professional world, whether companies or freelancers, all seeing it as a way to boost their productivity. And yet, AI is only in its infancy and has evolved a lot in a few months! The possibilities seem to be multiplying, especially with the release of GPT-4, a version of the language model that powers the chatbot more reliably and more accurately, and plugin integration in ChatGPT (see our article). Productivity tools are already undergoing changes: Microsoft has announced the arrival of the Copilot assistant for its Microsoft 365 suite, Google is preparing to add an intelligent assistant to its Workplace suite, and even Canva continues to integrating more and more AI into its tools via the Canva Visual Suite.
AI is gradually creeping into the world of work and seems inevitably destined to eliminate many current jobs. According to a survey by ResumeBuilder, relayed by Fortune, 25% of companies have already replaced employees with ChatGPT – this is the case for the media CNet, for example. It is a real revolution for jobs in the tertiary sector, as industrialization had been for manual professions, which will particularly affect Europe and the United States. So, what changes to expect?
Employment and AI: a revolution for the tertiary sector
In a blog postBill Gates believes it to be the most revolutionary technological advancement since the graphical user interface (GUI) introduced in 1980. “The development of artificial intelligence is as fundamental as the creation of the microprocessor, the personal computer, the Internet and the mobile phone. It will change the way people work, learn, travel, heal and communicate with each other “, he says. For the first time in the history of technological innovations, it is the so-called “intellectual” professions that would be the most affected. This change has been studied by several specialists, who try to predict the impact that this technology could have on the labor sector in the years to come.
So OpenAI researchers, in collaboration with Open Research and the University of Pennsylvania, tried to determine which occupations were most at risk among a sample of 1,000 occupations, each of which was then broken down into multiple tasks. According to their study, it appears that AI could take care of at least 50% of a large part of the tasks identified for each job. This “aid” would above all have an impact on so-called jobs “high income”because the latter involve the worker performing tasks using software.
ChatGPT: the most threatened professions
“Our results indicate that about 80% of the US workforce could see at least 10% of their work tasks affected by the introduction of AIs, while about 19% of workers could see at least 50% of their assigned tasks”, says OpenAI. Of course, the writing and creative professions (writers, poets, lyricists, journalists, graphic designers, etc.) would be particularly affected, but not only. The professions of law (such as legal assistants), programming and those particularly technical (mathematicians, accountants, digital interface designer, etc.) would also be concerned. On the other hand, occupations that “are highly dependent on science and critical thinking skills” would be less affected by the strong arrival of AI. Likewise, some industries that rely heavily on labor should escape it. This would notably be the case for forestry, social assistance and the food industry. Ditto for cooks, house painters, masons, stonemasons and mechanics.
Our colleagues from Internet Journal had fun asking ChatGPT what jobs he would be able to perform while being as efficient as a human. Here are some examples from his response:
Real estate agent : “I can provide property information, arrange virtual tours and help with contract negotiations.”
Legal Assistant : “I can carry out legal research, draft documents and help with the preparation of cases”.
Librarian “I can organize and catalog resources, as well as respond to user inquiries.”
Community Manager : “I can interact with users on social networks, moderate comments and manage advertising campaigns”.
Bank or financial adviser: “I can analyze financial markets and economic data to provide investment recommendations based on predictive models”.
Copywriter: “I can generate advertising content, catchy titles and product descriptions using persuasion techniques”.
Developer: “I can quickly generate codes, websites and propose design elements” And “Thanks to my algorithmic creativity, I can create captivating game scenarios and levels”.
Online teacher: “I can provide instant answers and educational resources based on student needs.”
Journalist : “I can write press articles based on reliable sources and respecting journalistic standards”.
Touristic guide : “I can provide information on the sights, culture and attractions of an area, as well as recommendations on activities, restaurants and accommodation”.
Moderator: “I can quickly identify and remove inappropriate content on online platforms”.
Call center operator: “I can handle incoming and outgoing calls, answer questions and resolve customer issues.”
Programmer: “I master several programming languages and can write or debug code quickly”.
Receptionist: “I can automate and efficiently manage call handling, appointment scheduling and information delivery.”
Recruiter (HR): “I can analyze CVs, select candidates and facilitate job interviews”.
Scriptwriter : “I can generate story ideas, dialogues and plots based on my extensive database of films, TV series and literature”.
300 million jobs lost: a profound change in the market
But concretely, what will it give? Analysts at investment bank Goldman Sachs just released a report on the issue and, according to them, no less than 300 million jobs worldwide could be eliminated by generative AIs like ChatGPT. Around two-thirds of current jobs in the US and EU would be exposed to some degree of AI automation. Here again, we find in the front line the legal, administrative and customer support professions for the United States, and the executives and administrative professions in Europe.
Thus, 46% of administrative tasks, 44% of legal tasks and 37% of architectural and engineering tasks could be replaced by AI. The life sciences, physical and social sector is also expected to be strongly affected with 36% of replaceable tasks, compared to 25% for commercial and financial activities rounding out the top 5 with 35%. As said before, professions requiring physical activity, such as construction and maintenance, are less exposed, with 6% and 4% respectively. Also according to the analysis, the five countries most affected by this white-collar revolution should be Hong Kong, Israel, Japan, Sweden and the United States, unlike China, Nigeria, Vietnam, from Kenya and India.
Seen like this, the future seems quite pessimistic. However, the study foresees an increase in productivity and an acceleration of the global economy, with annual global GDP which could increase by up to 7% thanks to generative AI. The bank’s analysts believe that their development could create new businesses. “The displacement of workers due to automation has always been offset by the creation of new jobs and the emergence of new professions”, they explain. In particular, we could think of developing jobs intended to test chatbots with text – a bit like what was done when Bing Chat was released in beta. It remains to be seen whether the job creations will be sufficient to compensate for the jobs at risk. The report is nonetheless reassuring, stating that “although the impact of AI on the labor market is likely to be significant, most jobs and industries are only partially exposed to automation and are therefore more likely to be supplemented than replaced by AI ‘IA.“Thus, only 7% of jobs in the United States could be totally replaced by AI, while 63% of workers would have their tasks completed by artificial intelligence and 30% would be completely unaffected by automation.
Employment and AI: an upheaval that calls for the greatest caution
The revolutionary impact of generative AI on the world of work is likely to be such that we must move forward with the greatest caution, as Sam Altam, CEO of OpenAI rightly reminded us, all the more so when we sees that this type of AI sometimes goes completely out of control when it is simply a question of requesting information and that it is already used for cyberattacks… An opinion shared by more than a thousand experts, including Elon Musk and researchers from DeepMind – the AI division of Alphabet, the parent company of Google. These latter indeed co-signed a letter calling for the immediate interruption and for six months of the development of artificial intelligences more powerful than GPT-4, judging the deployment of artificial intelligence too fast. This interruption would make it possible to assess in advance whether their effects will be positive and their risks manageable.
An entirely legitimate concern and request. With the “race for AI” which seems to have won them over the past few months, tech companies, not wanting to be competitive, sometimes release their technologies too quickly, without taking the time to put in place the necessary safeguards. This was particularly visible during the presentation of Bard, Google’s chatbot, which made a monumental mistake during its first presentation, or with Microsoft and Bing – the chatbot multiplied the errors and moods when it came out, ranging to insulting Internet users (see our article). There is also the question of the carbon footprinte, this kind of AI consuming a lot, but also those of the legal framework and the protection of privacy. And, above all, it remains to be seen how far we are ready to delegate to artificial intelligence, including for complex questions such as the verdict of a court decision or the diagnosis of a disease.