Goldman sees significant economic upside from AI adoption

WhatsApp Group Join Now
Telegram Group Join Now

Goldman Sachs economists anticipate considerable economic benefits from generative AI, expecting a 15% cumulative rise in US labor productivity and GDP. This projection is based on the assumption that the capital stock will align with the expanded labor potential resulting from widespread AI adoption.

MIT economist Daron Acemoglu’s recent research suggests a more modest impact: a 0.7% rise in total factor productivity and a 1.1% GDP increase. The bank’s economists attribute the differences to varying assumptions about AI automation’s reach and its impact on workforce distribution.

Goldman Sachs predicts that 25% of work tasks will be automated by generative AI, while Acemoglu’s estimate is lower at 4.6%. This gap stems from Acemoglu’s view that only 19.9% of tasks are AI-exposed, and of those, only 23% are cost-effective to automate within the next ten years. Additionally, Acemoglu doesn’t consider potential labor reallocation or the emergence of new tasks, unlike Goldman Sachs, which factors these into their projections.

Goldman explained that these assumptions account for more than 80% of the difference in our estimates, while the rest is due to smaller variations in assumptions about cost savings and production potential for workers not affected by displacement.

They also expressed agreement with Acemoglu’s perspective that automating numerous AI-exposed tasks isn’t currently cost-effective and may not be within the next decade for many tasks.

Economists emphasized that as technology costs decrease rapidly, there’s a high chance of more widespread adoption and automation, leading to potential cost savings.

Goldman Sachs disagrees with Acemoglu’s view that labor reallocation and new task creation are unimportant. They argue that history demonstrates technology-driven resource reallocation and the expansion of the production frontier as key drivers of economic growth.

The economists anticipate that AI will boost output by increasing demand in labor-advantaged areas and by creating new opportunities previously considered unfeasible due to technology or economics.

So, they see Acemoglu’s less hopeful perspective as pointing out valid concerns that significant macroeconomic effects of GenAI might not happen soon “and could be more delayed than commonly thought.”

WhatsApp Group Join Now
Telegram Group Join Now

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top