U.S. Federal Reserve States that AI hasn't yet increased Human Productivity
In a recent paper, economists from the Federal Reserve and Goldman Sachs have identified key challenges for the widespread adoption of generative AI in the economy. Despite heavy investment, less than 30% of AI leaders report satisfaction with the return on such investments, revealing gaps in identifying suitable use cases and realizing value. Organizations face shortages in skilled professionals and difficulties in instilling sufficient AI literacy across teams, hindering scaling efforts. Ethical and societal concerns complicate deployment, and regulatory complexity demands careful navigation to scale AI safely and effectively.
However, the outlook for generative AI is promising. The paper predicts that digital native companies will lead the wave of future innovations in generative AI, spurring related innovations such as specialized LLMs for specific domains and "copilots" like Microsoft's Copilot product. Economists project that if responsibly deployed, AI could boost global GDP by nearly 15% by 2035, adding more than 1 percentage point growth per year, comparable to the transformative impact of the industrial revolution.
The rapid growth of generative AI has already given rise to further innovations like agentic AI and landmark AI models like Deepseek's R1. The technology has been instrumental in novel drug discoveries and has been helpful in simulations to understand the nature of the universe. A huge spike, starting in 2023, of companies citing AI within research and development contexts and in corporate earnings calls shows that AI's integration with corporate innovation may have already begun.
The biggest challenge with generative AI right now is getting people and businesses to use it. Most companies outside of tech and the scientific fields are yet to integrate it into their daily operations, with the exception of the finance industry. Industry surveys show that AI adoption is far higher within large firms than small ones.
The timeline for AI's effects on labor productivity and GDP growth in the U.S. is predicted to start showing in 2027 and will accelerate to a peak in the 2030s, according to Goldman Sachs economists. The Federal Reserve recently published a paper suggesting that generative AI is not a passing tech trend, but a potential macroeconomic force with revolutionary effects on labor productivity.
However, the Fed also warns of the risk of investing too quickly in infrastructure for anticipated demand in generative AI, as demand not growing as expected can lead to "disastrous consequences," similar to how railroad overexpansion in the 1800s led to an economic depression towards the end of the century. The road to realizing the potential of generative AI will be slow and fraught with risk, but the paper compares generative AI to innovations like the microscope, implying that it could have a similar revolutionary impact on various fields.
The paper identifies generative AI as a type of technology with the potential for long-lived effects on productivity growth, similar to general-purpose technologies like the electric dynamo or the computer. The exact timeline for when we can expect the wave of generative AI innovations to take place is not specified in the paper. However, the outlook for generative AI as a macroeconomic force is clear: it has the potential to reshape industries, boost economies, and revolutionize various fields, making it a technology worth watching closely in the coming years.
Artificial intelligence, a technology that has the potential for long-lived effects on productivity growth, is being closely watched by economists, with generative AI being identified as a key area of focus. Companies are yet to fully integrate generative AI into their daily operations, but the tech industry, along with digital native companies and the finance industry, are leading the way. According to the paper, if responsibly deployed, AI could boost global GDP by nearly 15% by 2035, making it a technology worth investing in despite the challenges. In the near future, specialized LLMs for specific domains and "copilots" like Microsoft's Copilot product are expected to spur further innovations in generative AI.
