Artificial intelligence (AI) has become part of our lives. The latest development, generative AI, has burst onto the scene even more rapidly than previous developments. Starting with ChatGPT in November 2022, by May 2023 new and more powerful forms have become available on almost every system.
McKinsey has issued a report on the economic potential of generative AI, including the risks of this new technology. The full report is at https://www.mckinsey.com/capabilities/mckinsey-digital/our-insights/the-economic-potential-of-generative-ai-the-next-productivity-frontier#. This is a summary of their findings.
Generative AI could add trillions of dollars in value to the global economy. It will have a significant impact across all industry sectors, some more than others. It has the potential to automate work activities that almost 60-70%nof employees do today. It has more impact on knowledge work - occupations with higher wages and educational requirements. The pace of transformation is accelerating, about a decade earlier than previous estimates. We will need to address the challenges now.
McKinsey identified 63 use cases of generative AI spanning 16 business functions. Total annual economic benefits will be between $2.6 and $4.4 trillion. When modeling over 2100 detailed work activities McKinsey estimated additional annual productivity increases of $6.1 to $7.9 trillion.
McKinsey describes four examples of generational AI's benefits:
- Customer operations: Improving customer experience and agent productivity. It increases resolution of issues and reduces time spent. Productivity and quality of service improved most among less experienced agents.
- Marketing and sales: Boosting personalization, content creation, and sales productivity. Efficient and effective content creation, enhanced use of data, and improved search engine optimization (SEO). However, concerns about intellectual property rights and bias require human oversight.
- Software engineering: Speeding developer work and coding.
- Research and development: Reduced time, improved simulation and testing.
There may be additional benefits in all these categories which were not estimated by McKinsey.
The industries most likely to see the greatest impact are high tech, retail and consumer packaged goods, and pharmaceuticals. These industries do need to consider factors that could affect their ability to use generative AI. These include external interference, adversarial attacks, regulation, data constraints, privacy, and the need for human control.
Generative AI has shortened estimates of the time that technology could achieve human-level performance. It is now estimated that 60-70% of work hours can be automated, as early as 2030 to 2060.Adoption of generative AI undoubtedly will be faster in developed countries.
Previous generations of automation mostly affected collecting and preserving data. Generative AI wil have its biggest impact on knowledge work. Specifically, decision making and collaboration activities will be much more automated, while data management and physical activity will see little change. This also applies to higher education levels, which will approach those with some or no college education.
Generative AI could propel higher productivity growth. For this to happen, labor hours must be redeployed effectively.
Because of the pace of development, the challenges must be addressed quickly if generative AI is to be used responsibly.