In a remarkable development, Goldman Sachs has recently highlighted the emergence of advanced generative artificial intelligence models created by several Chinese companies, such as DeepSeek, that come at a significantly lower cost than existing productsThis breakthrough could not only accelerate the adoption of AI but also enhance its contribution to global economic growth.

The implications of this advancement challenge the prevailing notion that high investment costs represent the greatest barrier to entry for the most powerful AI modelsJoseph Briggs, co-head of Goldman Sachs’ global economics team, noted in the report that although the specifics surrounding how Chinese researchers developed their AI technologies and the total costs remain somewhat ambiguous, the lower cost structure could facilitate faster development and dissemination of AI on a global scale.

Briggs explained that, "If the reduced costs help increase competition in the development of platforms and applications, this breakthrough could enhance the macroeconomic upside in the mid-term." He pointed out that limited adoption of AI technologies currently poses the most significant bottleneck in realizing productivity gains associated with AI, and increased competition is likely to expedite the construction of AI platforms and applications, thereby driving further adoption.

However, he cautioned that the short-term impacts on adoption may be limited since cost itself isn’t currently the primary barrier to widespread usageAccording to data from the U.SCensus Bureau, the greatest short-term hurdles reported by companies in the U.S. are insufficient understanding of AI capabilities and privacy concernsA mere 6% of U.S. companies reported using AI for regular production tasks, an increase from 4% at the end of 2023.

So, how will AI contribute to GDP growth?

Goldman Sachs’ economic team has projected that widespread adoption of generative AI could boost U.S. labor productivity by around 15% over the next decade, primarily through the automation of work tasks

Advertisements

This could translate to an approximate increase of $4.5 trillion in annual U.SGDP (in current dollars). The anticipated economic benefits are expected to initially favor hardware and infrastructure providers, followed by platform and application developers, eventually manifesting in enhanced productivity and efficiency across broader industries.

Furthermore, the team predicts that the cycle of AI investment in the U.S. will gradually weaken after reaching 2% of GDP as the computational costs of training AI models and running AI queries decreaseAs the rate of end-user adoption increases, investment in AI software is expected to grow steadily.

Despite growing concerns over Chinese advancements in AI leading to skepticism about the investment and technological leadership of existing companies, the team's view on AI's macroeconomic impact remains unchangedThey believe that the primary macroeconomic drivers will stem from increased productivity as businesses integrate AI-driven automation into their operations.

Heightened global competition could encourage cross-border collaboration or the reduction of regulatory barriers to foster the development and adoption of AI.

Simultaneously, Briggs remarked that the potential automation and productivity enhancements brought about by generative AI are likely to be broadly similar across various economies. "While we continue to expect that due to the United States' leadership in AI model development, the country will adopt AI more rapidly than others, the emergence of non-U.S. platforms and applications could expedite the adoption timelines in other regions," he stated.

How might AI elevate productivity?

The team's projections assume that adoption of generative AI technologies in the U.S. will begin reflecting in productivity data by 2027, potentially reaching peak impact by the early 2030sIn these projections, other developed markets and key emerging market countries are expected to lag behind the U.S. by several years

Advertisements

Briggs noted, "Recent reports from DeepSeek suggest that adoption might happen sooner than we initially anticipated."

Goldman Sachs' research still anticipates a rise in AI adoption in the mid-term, with Briggs indicating that the types of work tasks that generative AI can automate could save each employee thousands of dollars annuallyHe stated, "Given the immense potential cost savings from generative AI and the likelihood of very low marginal costs once application development is completed, we believe the question of adopting generative AI is more a matter of 'when' rather than 'if.'

Briggs also acknowledged that there are reasonable questions regarding how low-cost AI models will influence stakeholders within the AI ecosystemThe distribution of any profits will depend on market concentration, intellectual property, scalability, and the ultimate competitive landscapeIt may be premature to fully grasp the impact of these new models; however, if expensive hardware and computational power become less critical to achieving economic advantages, companies focused on building physical infrastructure may realize diminished overall gains.

Nonetheless, Briggs pointed out that questions surrounding growth distribution may be less relevant to the overall macroeconomic narrativeThe prospects for economic advancement do not hinge on precisely who benefits; the overall impact of China’s breakthroughs will likely be net positive.

Will the development of AI in China lead to reduced investments?

An important question arises regarding whether more efficient AI models would lead to a decrease in AI capital expendituresStock analysts based on common estimates predict that by the fourth quarter of 2025, AI-related capital expenditures will rise to $325 billionWhether this reduction would also dampen GDP growth remains to be seen.

Goldman Sachs’ research indicates that if cheaper models lead to a decline in AI capital expenditures, two factors may limit the economic impact in such a scenario

Advertisements

While firms currently report increasing their AI-related investments, the actual impact on official GDP figures has been quite limited thus farAnalysts at Goldman Sachs suggest that companies are unlikely to make significant adjustments to their capital allocations solely because of the recent developments from China.

Furthermore, while low-cost AI models may reduce the anticipated construction of AI infrastructure, it's also plausible that these advancements will spur existing AI firms to ramp up investments in order to maintain their competitive edgeFundamentally, if new developments stimulate competition and lower costs, they could catalyze the quicker development of AI platforms and applications.