By Gilles Mathy
Edited by Nico Herlett
In a historic move, the US National Science Foundation (NSF) unveiled its National Artificial Intelligence Research Resource (NAIRR) Task Force in mid-2021. Fast forward to January 2024, the NSF launched the NAIRR pilot programme including ten federal agencies and 25 top IT giants. This bold step is part of the US agenda to come to the forefront of the AI market and creates urgency in the debate surrounding the social implications of AI adoption.
AI-generated image.
The NAIRR
Significant involvement of the Chinese government in AI R&D and China’s recent breakthroughs in the field, namely China’s “Supermind”, have intensified the US-China technological competition. Although the US private sector has proved to be America’s greatest strength in AI R&D, financial governmental stimuli are required to help the US private sector compete with China.
Recommended by the US AI Commission, the NAIRR Task Force represents a hybrid approach where the US government and the private sector join forces to keep their country ahead of the competition.
Through the provision of computational, data, software, model, training and user support, the NAIRR pilot consists of a national infrastructure that links American AI researchers with the resources they require to carry out their work. The purpose of the NSF project is in clear accordance with Joe Biden's presidential order on AI, which aims to maintain the U.S. as the world's premier hub for AI innovation.
The NSF estimates that democratising AI R&D through the NAIRR would cost 2.6 billion dollars. This cost, which must be split between public and private entities, is more than just the cost of asserting leadership in AI. The NSF states that AI has the ability to spur further innovation and discovery.
NAIRR is thus expected to have a positive spillover effect in other fields of research by facilitating the accessibility to new powerful tools.
Benefits of AI
In his Nobel-prize-winning research, Robert M. Solow identified various factors contributing to economic growth, with technological advancement standing out as a key driver. His work revealed a strong link between technological progress and economic prosperity. The US push for expanding its expertise in AI through initiatives like NAIRR, is therefore poised to reap the rewards with potential boosts to its economy on the horizon.
In fact, a study based on twelve developed economies conducted by Accenture in 2017 forecasts that by 2035, AI could double annual global economic growth rates through an increase in labour productivity by up to 40 per cent. However, this will not be the case for all economies but a selected few.
In reality, according to McKinsey, early adopters – meaning the companies that drastically adopt AI tools earlier – will most likely benefit disproportionately. On the other hand, the delayed adopters, or non-adopters, would probably observe some economic downturn.
Moreover, the benefits of AI for businesses, and entire nations, depend not only on their speed of adoption but also on the quality and efficacy of the AI tools that are adopted.
In sum, by promoting research into AI with the NAIRR Task Force, the US may be able to provide its flagship companies with cutting-edge tools before other countries can. Equipped with powerful software, American firms may gain competitiveness through increases in market share.
They may assert themselves as clear front-runners in their respective markets, thus widening the gap with foreign flagship companies.
Provided that the US champions the AI technologies, they may likely enjoy higher growth rates than in previous years. On the other hand, other countries risk observing lower or even negative growth rates if U.S. firms manage to dominate various markets.
AI-Caused Disruptions
Even though the NAIRR offers a great opportunity for the United States to prosper economically, such a project does not come without drawbacks. Indeed, AI is useful, but tends to create fear and labour market disruptions.
According to a quarterly CNBC survey of 2019, 24 per cent of all American workers surveyed fear their current job being eliminated within the next five years due to AI. Furthermore, the survey found that such fears grow with lower incomes.
The results show that 34 per cent of workers earning less than 50,000 dollars annually are afraid of losing their jobs due to technology. The number goes down to 16 per cent for workers earning between 100,000 and 150,000 dollars.
A McKinsey report, published in June 2023, suggests that the feeling that low-income households are more at risk than richer households is justified. In their study, lower-wage workers are up to fourteen times more likely to need to change occupations than those in highest-wage positions.
The explanation for that is that AI disrupts labour markets through job polarisation. Artificial Intelligence and automation have the potential to replace low-paying professions, normally requiring routine cognitive and manual abilities; but may also increase demand for well-paid skilled jobs that traditionally require non-routine cognitive skills.
Based on past trends, it could be argued that in the short and medium run, job loss will outpace job creation, while job creation will triumph in the long run.
As a result, AI's disruptive effects in the short and medium term could have an impact on economic inequality, income distribution, and salaries. Highly skilled labour which is able to use AI may observe an increase in pay due to rising demand, whereas many others may experience downward wage pressure or even unemployment.
Getting Through the Shift
With their NAIRR Task Force, the United States may take the lead in the competition for AI dominance. Although this seat has the potential to greatly boost the nation's economy, it also has the capacity to seriously destabilize the labour market and widen socioeconomic inequalities.
With the possibility to establish its domestic companies as front-runners in their respective markets, the U.S. needs to address the trade-off between enjoying great economic growth and increasing inequalities.
For the moment, the White House only implements policies pertaining to the development and application of AI that are reliable, safe, and secure.
Meanwhile, some experts have investigated and expressed the need for action regarding the socioeconomic impact of AI.
Bill Gates, for instance, formulated the idea of a ‘robot tax’ to compensate individuals for the jobs AI is replacing. There may also be a need for training programs to avoid rising unemployment.
Overall, policymakers must dig deeper into the subject of AI and its impact on the working class to avoid enjoying economic upturns at the expense of societal well-being.
Sources: The White House, McKinsey, National Science Foundation, CNBC, Accenture, MIT Press, MIT News, European Parliament
Written by Gilles Mathy
March 2024