Experts at the global investment bank Goldman Sachs have warned that generative AI threatens around 300 million jobs in developed economies. The Financial Times reports this.
Researchers suggest that around two-thirds of jobs in the US and Europe are automatable. In 63% of them, fewer than half of the tasks could be replaced by algorithms.
Analysts say that this category of professions still has time to adapt to changes or retrain for other forms of employment without significant disruption.
Around 30% of occupations are classified as physical labour and outdoor work. Analysts forecast the least impact from large language models (LLMs) on these professions. However, their tasks could be automated in another form.
In the remaining 7% of jobs, generative AI is capable of replacing more than 50% of tasks. Such workers could be fully displaced by computers.
Analysts say that lawyers and administrative staff are most at risk in the first instance. However, according to the report, those laid off will be able to find new jobs created by technological changes.
Analysts reached this conclusion citing economist David Autor’s article. According to them, today six in ten workers are employed in occupations that did not exist in 1940.
Among candidates for new positions could be police officers or cybersecurity professionals.
The forecast assumes that companies will continue investing in AI. Analysts compared the current trajectory of large language models with the growth of the IT sector in the 1990s.
If these trends align, spending on deploying LLMs in the United States could reach 1% of GDP by 2030, amounting to about $230 billion, based on current data.
Analysts also forecast that, thanks to the use of large language models, global labour productivity will rise. As a result, global GDP will grow by about 7% over the next ten years.
In March, Europol warned about the danger of criminals using ChatGPT.
In the same month, OpenAI researchers predicted that GPT models would affect 80% of jobs in the United States.