Goldman Sachs CEO Solomon: Compute Demand Far Exceeds Supply, AI Financing Will Remain Strong Over the Next 12 to 24 Months

Cailian Press Financial News Agency

Open the APP

Goldman Sachs CEO speaks out: Compute demand far exceeds supply; AI financing will remain strong over the next two years

Cailian Press, Shi Zhengcheng, Editor: Lin Qi, 2026-11-03 16:32

① Goldman Sachs Chairman and Chief Executive Officer David Solomon said that despite increased market volatility, investor demand to subscribe to large AI financing deals remains "very strong," and there will be strong financing over the next 12 to 24 months to meet part of the compute demand; ② He believes that AI infrastructure construction is still in its early stages; although bumps are inevitable ahead, looking out five to ten years, the potential of AI applications and the productivity gains they drive still excites him.

Cailian Press, November 3 (Editor: Shi Zhengcheng) — As the market continues to ask whether the AI investment boom can last, the Wall Street bankers raising funds for tech giants have offered a judgment from the front lines of financing: demand remains strong, and capital markets are still willing to support this round of construction.

On November 3, Goldman Sachs Chairman and Chief Executive Officer David Solomon said while attending the International Financial Leaders Investment Summit in Hong Kong, China, that despite increased short-term market volatility, investors remain highly focused on the growth of artificial intelligence and compute demand, and demand to subscribe to large financing deals is "very strong."

"One thing is quite clear right now: compute demand far exceeds supply," Solomon said. "Over the next 12 to 24 months, there will be strong financing activity to meet part of that demand."

Capital markets remain open

Solomon said Goldman Sachs was very busy last week, with major financings by SpaceX and Meta, as well as news related to Anthropic's upcoming IPO, all appearing in the same period. Despite some market volatility, capital markets remain active and willing to take on these deals.

As the world's top investment bank, Goldman Sachs, after leading SpaceX's "largest IPO in Earth's history" in June this year, also won the lead underwriter role for Anthropic's IPO. According to other reports, the bank also co-led Meta's $95 billion equity financing plan and served as the sole structuring agent for a $40 billion chip financing for SpaceX.

Solomon believes that before the direction of development changes, this tech super-cycle will continue to dominate capital markets and financing activity.

"Whether in Asia, the United States, or Europe, what is really driving capital markets and financing activity is the technology industry ecosystem," he said.

Solomon pointed out that bond investors first need to avoid losses, and when they believe the market environment has become more challenging, they will demand higher returns; equity investors are willing to take on more risk in pursuit of upside from technological development and commercial applications.

Financing costs determine the economics of projects

Solomon believes that although today's deals and tech companies are larger than in previous technology cycles, global capital markets have also expanded significantly. He said the total market value of globally listed equities is about $160 trillion, and there is a large amount of capital available to support growth.

In his view, the related discussion sometimes conflates two issues: whether there is money in the market, and at what price investors are willing to take on risk.

"I don't think this affects the availability of capital; I think it affects the cost," Solomon said.

Changes in the cost of capital will directly affect the economics of projects. He pointed out that if financing costs rise too much, the returns of some projects will change, which in turn will affect how projects proceed and what financing methods are used.

Construction is still in its early stages, but it will not rise in a straight line

Technological evolution may also change the construction path. He pointed out that new technologies and advances in existing technologies may change the trajectory of infrastructure construction as a whole, so it is difficult now to accurately judge how much money the world will ultimately invest.

Morgan Stanley Chief Executive Officer Ted Pick, who also attended an event in Hong Kong, China, the same day, likewise believes this investment cycle is still in its early stages.

According to an earlier Cailian Press report, Pick estimated that large tech companies may invest $800 billion to $1 trillion per year over the next 7 to 12 years; considering the entire construction cycle, a cumulative investment scale of $10 trillion to $15 trillion is not an unreasonable assumption.

Using a baseball game as a metaphor, he said it is roughly at "the second inning of a nine-inning game." Pick also expects that the related productivity gains could contribute about 0.5 to 0.7 percentage points to U.S. real GDP growth per year.

Post-election focus on 2028; AI governance needs clear responsibility

Turning to the U.S. midterm elections, Solomon did not predict the outcome, but said that regardless of the result, discussion will soon turn to the 2028 presidential election.

He said changes in congressional control will of course draw attention, but from a substantive policy perspective, what is more worth watching is what changes may occur in the next administration and how the related outlook evolves over the next two years.

He laid out three basic principles: governments need to set rules; technology developers need to be responsible for how products are developed and brought to market; and companies that use technology, such as Goldman Sachs, must be responsible for how technology is deployed in their own businesses.

Citing Goldman Sachs' use of an AI trading platform for 20 years, Solomon pointed out that companies cannot simply shift responsibility to technology after a system causes losses. Even if models can complete work more autonomously, people must still bear responsibility for technology applications, with effective risk control and governance mechanisms in place.

At the same time, he stressed that Goldman Sachs has about 45,000 employees and needs to properly manage access to tools; not all employees can use every tool at any time.

On model selection, Solomon said Goldman Sachs maintains partnerships with several major AI vendors. He noted that some uses, such as cybersecurity, require frontier models, but much of the work to improve employee efficiency and transform business processes does not require the most advanced model capabilities.

As for open-weight models, he acknowledged that their spread is expected to lower costs, but stressed that given strict governance requirements, Goldman Sachs will proceed prudently and does not need to be the earliest deployer.

Global Market Intelligence

US Stock Market News

Cailian Press statement: The content of this article is for reference only and does not constitute investment advice. Investors who act on it do so at their own risk.

Business Cooperation

Popular Unlocks

[Trend Research Report · Industry] AI short dramas and AI comic dramas are the areas where AI applications are scaling up first; they have already generated significant revenue contributions + this year's market size is expected to reach 40 billion yuan. Analysts sort out leading companies with high-quality IP, content production capabilities and distribution resources; this company holds scarce refined oil export quotas

① AI short dramas and AI comic dramas are the areas where AI applications are scaling up first; they have already generated significant revenue contributions + this year's market size is expected to reach 40 billion yuan. Analysts sort out leading companies with high-quality IP, content production capabilities and distribution resources; ② Since the third quarter, overseas refined oil price spreads have remained high, and this company holds scarce export quotas, with net profit attributable to parent company up 800% year on year in the first three quarters.

[Intraday Treasure] As policy planning details gradually land, institutions say order certainty in this industry may improve, and multiple sub-sectors are expected to usher in high-growth development. This enterprise has the R&D and production capacity for multiple models of products

As policy planning details gradually land, institutions say order certainty in this industry may improve, and multiple sub-sectors are expected to usher in high-growth development. This enterprise has the R&D and production capacity for multiple models of products, and another enterprise has completed delivery of a sub-segment product, with business scale hitting a record high.

[Intraday Treasure] Bookings are earlier and growth is faster this year; as the peak season arrives, this sector continues to heat up. This company has relevant resource advantages

Customers are entering earlier, booking volume has doubled year on year, and as the peak season arrives, this industry continues to heat up. This company has relevant resource advantages.

Goldman Sachs CEO Solomon: Compute Demand Far Exceeds Supply, AI Financing Will Remain Strong Over the Next 12 to 24 Months

Topics

Tags

Body · AI translation