The Economic Potential of Generative AI: The Next Frontier of Productivity

McKinsey's latest report points out that the development speed and potential impact of generative AI far exceed expectations. The report predicts that AI may reach human-level capabilities sooner than imagined, with a median forecast before 2030. Compared to the predictions made in 2017, the new report is more optimistic about the prospects of AI development.

Generative AI has penetrated all aspects of life. Unlike early AI, which was limited to specific fields, such as AlphaGo's achievements in Go, current generative AI tools like ChatGPT and Stable Diffusion are widely influencing people's daily lives and work. The popularity of these tools enables everyone to utilize AI for various tasks such as creation and drawing.

The report highlights the astonishing speed of AI development. In just a few months, AI capabilities have made a qualitative leap. For example, ChatGPT powered by GPT-4 significantly outperforms GPT-3.5, and Anthropic's Claude can now process about 100,000 tokens per minute, which is ten times what it was a few months ago.

McKinsey's analysis adopts two complementary perspectives to assess the economic value of generative AI:

1. Analysis of enterprise use cases: 63 generative AI use cases have been identified, covering 16 business functions. If widely applied across various industries, it could bring an economic benefit of $2.6 to $4.4 trillion annually. This is 15%-40% higher than the predictions made in 2017.

2. Analysis of the Impact on Professions: An assessment of the potential impact of generative AI on approximately 850 professions. Considering the overlap with the first perspective, the total economic benefit of generative AI is expected to reach between $6.1 trillion and $7.9 trillion annually.

From the perspective of the technical impact on the proportion of functional costs, customer operations, marketing and sales, software engineering, and R&D account for about 75% of the total value of generative AI use cases. In contrast, the potential value in fields such as manufacturing and supply chain is relatively lower.

Generative AI may also bring additional value to the entire company by improving enterprise knowledge management systems. Its powerful natural language processing capabilities can help employees retrieve internal knowledge more conveniently, thereby enhancing decision-making efficiency.

The report predicts that, based on current performance, generative AI will reach human-level capabilities faster than previously estimated in various aspects. For example, the median time for AI to achieve human-level natural language understanding has been moved up from 2027 to 2023.

Experts predict that generative AI will have the greatest impact on knowledge work, particularly in decision-making and collaboration. The potential for automating specialized knowledge has increased by 34 percentage points, and the potential for automating management and talent development has risen from 16% in 2017 to 49% in 2023.

In the face of the opportunities and challenges brought by the rapid development of AI, the report calls on all parties to actively respond:

• Business leaders need to consider how to leverage AI to create value, manage risks, and adjust human resource strategies.
• Government decision-makers should pay attention to the impact of AI on labor planning and formulate corresponding policies for support.
• Individuals need to remain vigilant while enjoying the conveniences of AI and actively participate in related decision-making.

Overall, this report provides a comprehensive analysis of the profound impact of the generative AI boom on the socio-economy, offering important references for various sectors to cope with future changes.