The personal computer; internet; cloud computing and mobile computing have unmistakably changed most parts of the economy; education and learning; work and leisure pursuits. It seems likely artificial intelligence also will do so.
But how much impact AI might have on outcomes or productivity is open to question. Studies of earlier computing technologies (PCs, cloud computing, internet) on outcomes and productivity gains have found uneven impact.
Researchers and analysts still debate the degree of quantifiable impact (both positive and negative), even if the impact on life, work and learning might seem obvious.
Study | Technology | Key Findings |
Brynjolfsson and McAfee (2014) | IT and productivity | While IT has led to productivity growth, the distribution of benefits has been uneven, with some industries and workers experiencing greater gains than others. |
Autor et al. (2013) | Computerization and jobs | Computerization has led to job polarization, with a decline in middle-skill jobs and growth in high-skill and low-skill jobs. |
McKinsey Global Institute (2017) | Automation and jobs | Automation could displace up to 800 million jobs globally by 2030, but it could also create new jobs and boost productivity. |
Davenport and Harris (2019) | AI and productivity | AI has the potential to significantly increase productivity, but its benefits will depend on factors such as organizational culture, talent, and data quality. |
Forrester (2019) | AI | AI is expected to create $5.8 trillion in economic value by 2022. |
World Economic Forum (2016) | Fourth Industrial Revolution | The convergence of technologies, including AI, IoT, and robotics, is reshaping the labor market and requiring new skills. |
Hernández-Murillo (2003) | Computers | Benefits from computer use persist long after investment, with gains in TFP growth from 1995-99 computer investment expected to peak around 2006. |
McGuckin et al. (1998) | Computers | Computer-intensive manufacturing sectors saw labor productivity growth jump to 5.7% annually in 1990-1996, compared to 2.6% in other sectors. |
McKinsey Global Institute (2024) | AI | Generative AI could potentially add more than 0.5 percentage points to productivity growth. |
Ntiva analysis, 2020 | Cloud computing | Cloud computing improves employee productivity through enhanced collaboration, reduced downtime, improved data management, and facilitating remote work. |
Unnamed study (EconStor), 2022 | Cloud computing | Cloud adoption significantly improves labor productivity for firms in manufacturing and information/communication services sectors. No impact on IT investment found across sectors. |
Unnamed study (NCBI), 2023 | Cloud computing | Cloud computing integration positively impacts financial, environmental, and social performance of SMEs. Complexity, cost reduction, and government support are top factors influencing cloud adoption. |
Internet Access and its Implications for Productivity, Inequality, and Resilience (2021) | Internet | Universal access to high-quality home internet service would raise earnings-weighted productivity in the post-pandemic economy by 1.1%, implying flow GDP gains of $160 billion per year. |
The Impact of the Internet on Industrial Green Productivity: Evidence from China (2022) | Internet | The use of the Internet is conducive to both a reduction in energy intensity and an improvement in energy efficiency in industrial sectors. |
The Economy and the Internet: What Lies Ahead? (2001) | Internet | Even a few tenths of a percent impact on productivity growth rate from the Internet could represent a significant portion of any permanent surge in productivity. |
Mobile and more productive? Firm-level evidence on the productivity effects of mobile internet use (2016) | Internet | The study found evidence of productivity effects from mobile internet use at the firm level, though specific figures are not provided in the search results. |
Internet Access and its Implications for Productivity, Inequality, and Resilience (2021) | Internet | Universal access to high-quality home internet service would raise earnings-weighted productivity in the post-pandemic economy by 1.1%, implying flow GDP gains of $160 billion per year. |
But most of us would likely agree that the positive benefits have been greater in some industries than others, suggesting that AI should also have disparate impact. Looking at computing technologies in general, some would say there are many industries where the actual productivity impact of applied computing technology has been relatively muted.
That is not to say computing has had no impact, but simply that business outcomes have been varied. Almost all higher-order machines use computing to some extent, as do most job functions. But applied computing often is not a key driver of business results.
High Impact Industries | Low Impact Industries |
Information Technology | Agriculture |
Financial Services | Construction |
E-commerce | Healthcare |
Telecommunications | Education |
Media & Entertainment | Hospitality |
Automotive | Mining |
Manufacturing | Forestry |
Aerospace | Fishing |
Logistics | Textiles |
Energy | Artisanal Crafts |
AI might also have disparate impact on job functions. How “disruptive” the impact might be also is open to question. For marketing functions, for example, it is not entirely clear that AI radically changes possibilities, other than to make all automated processes even more precise, the costs of doing so lower and the effectiveness of current tools higher.
AI should make today’s personalization efforts even more precise; predictive analytics possibly more accurate; automated decisionmaking more prevalent; content creation and customer service more automated and many operations more efficient.
But those trends already are in place. AI enhances them, but might otherwise have less “disruptive” impact than some believe.
No comments:
Post a Comment