Despite the onset of personal computers, the internet, and other revolutionary technology, many developed nations are struggling with stunted economic growth. Projection for the Organisation for Economic Co-operation and Development (OECD) nations suggest an overall expansion of only 1.7% for the current year, an anomaly often referred to as the productivity paradox by economists.
Artificial intelligence (AI) is touted as the game-changer to break this pattern of mediocrity, yet scepticism looms. One noteworthy paper by Daron Acemoglu, a labour economist from the Massachusetts Institute of Technology, has ignited fervent discussion.
Acemoglu’s paper posits AI’s potential to enhance worker performance is only ‘modest’ and that it will contribute to US economic output less than 1% over the coming ten years. This contribution is meagre when compared to predictions made by economists at Goldman Sachs last year who estimated that generative AI could boost global GDP by 7% within the same timeframe.
Those on the bullish end of the spectrum, including Sam Altman of ChatGPT maker OpenAI and Jensen Huang of Nvidia, believe AI holds massive potential, with ambitious goals like eradicating poverty and igniting ‘the next industrial revolution’. However, if these optimists prove incorrect, this could spell trouble for advanced economies desperately in need of productivity improvements as their workforce ages and begins to decrease.
Regardless of widespread optimism, Acemoglu remains steadfast in his belief that AI won’t reverse stagnation as presented in an interview with The New York Times. He believes the existing technology can automate merely about 5% of an office worker’s jobs. He further asserts that while AI has much more to offer to tackle the productivity issue, it currently isn’t enabling this, leading to his concern over the prevailing hype.
According to Acemoglu, AI has the potential to automate routine tasks fast and efficiently, freeing up humans to engage in intellectually demanding tasks such as devising a business strategy for launching a new product. However, he remains dubious about the technology’s ability to aid workers in improving their problem-solving skills or taking on intricate tasks. He insists that achieving this would lead to a significant increase in worker productivity.
AI holds the power to transform professional livelihoods on par with the effect of the plough, according to some industry experts. However, not everyone agrees with this optimistic view. A long-time commentator on the impact of technology on the labour market and economy, Acemoglu, has expressed concerns that the increasingly competitive race among big tech companies to lead in AI could bring societal instability.
Others, too, have joined him in showing scepticism towards the hype surrounding AI, including David Cahn of Sequoia, a leading venture capital firm, and teams at Barclays and Goldman Sachs. These experts suggest that companies’ growing investment in AI could lead to a speculative bubble. Interestingly, Sequoia has recently backed Fireworks AI, a start-up.
Contrary to Acemoglu’s belief, detractors assert that his perspective underestimates AI’s potential to drive scientific progress and disrupt the corporate sector. Economist Tyler Cowen, for instance, argues that AI will create most of its value by eliminating the least effective companies, outlining a flaw in Acemoglu’s models.
Lynda Gratton, a London Business School management practice professor, who consults on AI adoption, remains hopeful but notes that it might be premature to label AI as a productivity booster until its effects are observed in real-world workplaces in the upcoming years. She acknowledges the testing of AI in corporations presently.
She emphasises the role of AI as a tool able to reduce expenses and increase efficiency for many businesses. However, firms focusing on growth need innovation-focused tools, she notes.
Similarly, Acemoglu speaks to the need for fostering innovation to enhance workplace productivity and competitiveness, considering the ageing population. He believes if AI can automate approximately 40 per cent of an average worker’s tasks; he might have to reconsider his stance on the technology, insisting that he is “not a complete pessimist”.
This article was initially published by The New York Times.