Goldman Sees $1.5 Trillion Flowing To AI. Here’s Where To Invest

The economists at Goldman Sachs have concluded that more than a trillion dollars could be invested in generative artificial intelligence over the next decade, according to Goldman Sachs Economics Research.

To get your share of that money, you can start a company, invest in a startup, or buy stock in a publicly-traded company likely to benefit most from Generative AI.

Companies likely to enjoy higher revenues from Generative AI — such as Nvidia and Microsoft MSFT — are pretty well known. Their stock prices could already reflect optimistic assumptions about how much Generative AI will improve their business.

There could be better investment opportunities in less well-known stages of the Generative AI ecosystem. For example, investors may wish to consider shares of companies — such as Accenture ACN , ServiceNow NOW , and Snowflake — which have introduced Generative AI-related products and services — but have not quantified their future revenues.

This uncertainty could provide considerable upside if Generative AI helps the companies grow faster than investors expect.

Goldman Sachs Sees Significant Investment In Generative AI

Significant capital will flow into Generative AI. According to Goldman Sachs Economics Research, by 2025, the capital investment could total $200 billion. By 2033 — combining Goldman Sachs’ assumptions about Generative AI investment as a percent of GDP and a Congressional Budget Office GDP forecast — I estimate Generative AI investment in the U.S. could top $1.5 trillion by 2033.

Goldman wrote it “looked at previous world-changing innovations in electricity and personal computers” which sparked investment as high as “2% of U.S. GDP as the technologies were adopted into the broader economy.”

By enhancing worker productivity, Goldman sees Generative AI eventually having an even bigger impact. As Joseph Briggs, a Goldman Sachs economist, explained in an August 7 interview, “The return on investment in Generative AI will come mostly from freeing up the time of workers to perform more productive activity, and less from displacing workers.”

Generative AI investment could peak at a range between $980 billion to $1.57 trillion by 2033. How so? According to the Congressional Budget Office, U.S. GDP is forecast to reach $39.2 trillion by 2033. Goldman Sachs envisions AI-related investment could to peak “as high as 2.5 to 4% of GDP in the U.S.” in the next decade.

Why does Goldman see Generative AI as having the world-changing potential of the PC or the electric motor? As Briggs told me, there are three reasons:

• It’s generalized. While AI has been around for decades and was used for specialized domains, only recently has AI been generalized.
• It’s original. GAI can produce original results — in the form of text, images, and multimedia — that are indistinguishable from human output.
• It’s easy to use. Finally, in a big change from last year, Generative AI is approachable using natural language. Non-technical employees no longer need data scientists to mediate their information requests.

It will take time for the investment in Generative AI to reach its peak as it flows from technology providers to corporate end-users. “By 2025, some $200 billion will be invested in Generative AI. First the money will have a direct benefit for hyperscalers — including cloud services providers and Nvidia — who are developing the technology. After that initial investment has been made, Generative AI will be widely deployed throughout the economy which will boost global productivity by more than one percentage point a year in the decade following widespread usage,” he said.

Specific industries are likely to be early adopters. As Briggs said, “We also looked at business surveys and studied how transformative technologies such as the PC and electric motors were deployed. We don’t expect the biggest changes to be on the next three years — it is in the three to 10 year period after the initial launch — between 2025 and 2030 — that we will see a significant impact. In the initial period, larger firms in industries — such as information, professional, scientific, and educational — that have already adopted AI are most likely to adopt Generative AI first.”

Goldman sees ChatGPT as a catalyst for conversations between companies and investors. “Market interest in AI has already increased rapidly, with more than 16% of companies in the Russell 3000 mentioning the technology on earnings calls, up from less than just 1% of those firms in 2016,” Briggs wrote.

Companies With Quantified Generative AI Revenue Forecasts

Since May 24, when Nvidia published its eye-popping revenue forecast for the current quarter, investors have bid up shares of the most obvious beneficiaries of GAI investment.

Two companies whose shares have soared since then may already reflect much of that upside in their share prices:

Nvidia Stock Up 48% Since May 24

Nvidia’s GPU chips and its software and services give it a leadership position in helping companies use Generative AI. On May 25, Nvidia reported on its April 2023-ending first quarter — including a Q2 revenue forecast of $11 billion — 54% above estimates.

Nvidia’s stock has risen considerably and is highly valued. At a price/earnings ratio of 57.2, Nvidia is quite expensive compared to rivals’ average P/E of 21.5, according to CNBC. Nvidia’s 12-month price target of $504.46 is some 11% higher than its August 7 price of $454.17, TipRanks noted.

Microsoft Stock Up 4.5% Since May 24

In January, Microsoft invested $10 billion in OpenAI, the developer of ChatGPT. Moreover, Microsoft plans to launch CoPilot, a ChatGPT fueled assistant for users of Excel, PowerPoint, and Word I estimated in April could add $40 billion to the company’s revenue. In July, Macquarie forecast CoPilot could add $14 billion to Microsoft’s revenue.

Microsoft stock has not risen much since May and its valuation is reasonable. At a price/earnings ratio of 30, Microsoft is slightly more highly-valued than its industry peers — which trade at an average P/E of 28.2, according to CNBC. Microsoft’s 12-month price target of $392.97 is about 19% above its August 7 price of of $330.11, TipRanks noted.

Companies With Uncertain GAI Revenue

Investors could benefit from buying stock in companies which have not made specific forecasts of how much revenue AI will add to their top lines. This bet could pay off well if Wall Street is underestimating how much Generative AI will add to their revenue.

Accenture Stock Up 9% Since May 24

Companies will hire consultants to help them evaluate whether and how to use Generative AI. Accenture — while not providing an estimate of how much additional revenue it will generate — has invested $3 billion and is hiring 40,000 people — doubling its staff to help clients with AI, according to the company.

Accenture stock has increased modetately while its valuation is a bit elevated. At a price/earnings ratio of 27.2, Accenture trades above its industry peers’ average P/E of 23.9, according to CNBC. Accenture’s 12-month price target of $344.55 represents a 10.29% upside opportunity from its August 7 price of $312.40, TipRanks noted.

ServiceNow Stock Up 9% Since May 24

ServiceNow is introducing AI-powered services and has declined to quantify how much they will add to its revenue. For example, as CEO Bill McDermott told me in a July 26 interview, ServiceNow is developing AI Lighthouse — in collaboration with Nvidia — which will use Generative AI to develop services to boost productivity for specific industries.

ServiceNow stock has risen modestly while its valuation is high compared to peers. At a price/earnings ratio of 55, ServiceNow is valued much more highly than its industry peers’ average P/E of 39.2, according to CNBC. ServiceNow’s 12-month price target of $638.24 represents 16.76% upside from its August 7 price of $546.62, according to TipRanks.

Snowflake SNOW Stock Down 11.3% Since May 24

Snowflake is developing new products and has formed a partnership to help companies benefit from large language models (LLMs), according to my May 2023 interview with CEO Frank Slootman. While declining to discuss how much additional revenue these would add, a spokesperson told me the generative AI market is worth “hundreds of billions of dollars.”

Snowflake shares have fallen and its valuation is very high compared to peers. At a price/earnings ratio of 273.2, Snowflake is much more highly valued than its industry peers’ average P/E of 69.5, according to CNBC. Snowflake’s 12-month price target of $190.27 is 21.08% above its August 7 price of $157.15, TipRanks noted.

If you are patient and willing to bet these three companies will win a significant piece of the Generative AI pie, you could invest in their shares and wait until they make that happen.