Generative AI — AI buildout continues as monetization challenges emerge

In September 2023, we published a report titled “The ascent of generative AI — What investors should know”. Since then, there have been a number of significant developments in the AI environment. Now that we are revisiting the topic a year later, we felt it would be beneficial to review what has changed, what still stands, and the key themes we expect to impact investors.

1. Transition from initial excitement to implementation and rationalization

We believe generative AI has likely progressed from the “peak of inflated expectations” stage in the Gartner Hype Cycle into the “trough of disillusionment” stage. Although interest has not waned, we believe that management teams continue to balance expenses with productivity gains while also planning implementation on a larger scale.

2. Release of OpenAI model series capable of complex reasoning

The new OpenAI o1 series of large language models are designed to spend more time thinking through harder problems in science, coding, and mathematics before they respond, and they require a significant amount of inference computing resources. Another member of the OpenAI o1 series includes a smaller and more cost-effective version of the model.

3. Emphasis on infrastructure, power, and cooling needs of AI data centers

Following more than a year of excitement, the attention around AI has expanded beyond just technology firms — it now also includes those building out the data-center infrastructure and supporting the need for reliable power and efficient cooling. We expect ongoing interest in generative AI to continue driving spending to build out and support data centers.

4. Acute focus on profitability and sustainability as applications have expanded

We continue to believe that the enterprise environment is key for generative AI to grow to scale. The focus for enterprises has now shifted to returns on investment for the capital being allocated to developing and implementing these new AI-capable software platforms, and we see opportunities for companies able to deliver on this front.

AI has been a dominant theme in the stock market over the past few years with the potential to have much the same effect on the economy. Generative AI has the ability to create new content, moving beyond traditional AI’s data analysis and improved forecasting capabilities. It is the latest and, perhaps, most important step so far in the technology’s evolution, preceded by traditional AI systems, machine learning, deep learning, neural networks, and automated speech recognition and language processing.¹

In the past, traditional AI has been helpful in predicting outcomes and focusing on a narrow task, such as accurately identifying and categorizing images. Generative AI, on the other hand, is based on a text prompt, whereby a user types text-based inputs in the form of a question or request, and the large language model does its best to produce an appropriate response.² It offers broader applications, using deep-learning techniques and large language models to generate unique and new content using trained data. A few examples of this advancement include images, videos, music, essays, language, conversations, and even software code itself. Multimodal large language models represent another example of advancement— unlike large language models based on generative AI that handle text inputs and outputs, these models process information from multiple types of media, including images, videos, audio, and text.

AI has been around since the 1950s and has experienced several false starts along the way. However, over the past two decades, the convergence of big data, advances in high-performance computing power, advances in machine-learning algorithms, and cloud computing were instrumental in driving recent AI advancements. In terms of size and importance, it is widely believed that generative AI represents the next major computing platform shift where various types of software and applications will be developed. However, it may take several years for these generative-AI applications to develop, and demand may not materialize until the buildout phase of the core AI-focused infrastructure has been completed.

¹ Deep learning refers to a type of AI that uses layers of neural networks to mimic how humans learn.

² A large language model is a deep-learning neural network that is trained on massive amounts of existing text to generate new text similar to human language.

According to Bloomberg and International Data Corporation (IDC), the overall market for generative AI could grow at a compound annual growth rate of approximately 49%, from $11 billion in 2020 to $1.36 trillion in 2032 (Chart 1). The rapid growth in generative-AI tools has been driving demand for semiconductor chips, software, networking equipment, memory, and storage solutions.

Chart 1: Generative AI market Sources: Wells Fargo Investment Institute, Bloomberg, and IDC. Data as of March 28, 2024. E = estimated. Estimates based on Bloomberg Intelligence forecasts using IDC, eMarketer and Statista data for the overall generative AI market.

Over the past few years, a considerable part of generative AI’s initial infrastructure-buildout phase has been driven by the need to train the latest large language models on increasingly large datasets of text and code. The costs associated with developing, training, and managing generative-AI large language models effectively are fairly prohibitive as these models are very compute-, semiconductor-, networking-, and storage-intensive. Further, the high cost of training large language models can increase exponentially as models grow in size, which occurs as they are trained on more parameters (that is, variables in a neural network³ that the model learns during the training process).

Although the vast majority of the capital expenditure (capex) spending cycle is being allocated toward training large language models, inferencing (the thinking that occurs when the already trained large language model generates a response to an inquiry) is also gaining traction. OpenAI’s new o1 series of models are an example of this — the models are capable of complex reasoning and will require higher levels of computing resources for both training and inference, but particularly for inference. According to Bloomberg and IDC, the overall market for training could grow at a compound annual growth rate of approximately 31%, from $32 billion in 2022 to $471 billion in 2032. Meanwhile, the overall market for inferencing is projected to grow at a compound annual growth rate of 49%, increasing from $3 billion in 2022 to $169 billion in 2032 (based on inferencing revenue; Chart 2).

Chart 2: Forecasts for training versus inference revenues Sources: Wells Fargo Investment Institute, Bloomberg, and IDC. Data as of March 28, 2024. E = estimated. Estimates based on Bloomberg Intelligence forecasts using IDC data.

We are also seeing more evidence of generative AI moving to the edge, meaning that generative AI models are being deployed directly onto local edge-computing devices (such as PCs and smartphones) that allow devices to process data locally instead of relying on the cloud. The PC original equipment manufacturers (OEMs) are optimistic about AI PCs — that is, PCs with generative AI integrated into both the hardware and software. Demand for these devices may drive an overall PC-replacement cycle in 2025. Looking back, as we emerged from the pandemic, consumers shifted their preferences from consumption of goods to services and experiences, which led to double-digit year-over-year declines in PC unit demand. Additionally, we expect more consumer-centric smartphones to be launched that can run on-device generative-AI models this year.

³ In deep learning (a subset of machine learning), a neural network is a mathematical model that can learn and model relationships between input and output data that are nonlinear.

Over the next two decades, we believe that AI will reach far beyond the software industry to transform the health care, education, manufacturing, transportation, financial, and retail industries, among others. In our view, generative AI has the potential to accelerate innovation, improve productivity in the workplace, and drive faster economic growth in the coming decade. Yet, historically, markets have often priced in technological advances well in advance of adoption. Additionally, there are several evolving risks that need to be understood in this case. The pace of adoption may be slower than the market is currently expecting. Debates among the investment community regarding when hyperscalers⁴ will generate an adequate return on investment in terms of sufficient generative-AI revenue monetization and the sustainability of the AI capex super cycle underway remain front of mind. The rise of generative AI could lead to significant white-collar job losses that could dampen economic growth and raise social tension at some point in the future. Investors also should note that government regulation so far has been uneven across global regions. That regulation is likely to accelerate and may need time to balance economic growth and social concerns.

We always favor a clear-eyed evaluation of risk and reward in every investment decision, and that remains true in the early stages of even a transformative technology. As we have previously noted, history teaches us that investors have tended to overestimate the impact of a new technology in the short run — even if they eventually underestimate its long-term impact.

⁴ Hyperscalers refers to hyperscale cloud providers, a small group of companies who operate large networks of large data centers to facilitate cloud computing.

All investments are subject to market risk, which means their value may fluctuate in response to general economic and market conditions, the prospects of individual companies, and industry sectors due to numerous factors, some of which may be unpredictable. Be sure you understand and are able to bear the associated market, liquidity, credit, yield fluctuation, and other risks involved in an investment in a particular strategy.

Equity securities are subject to market risk which means their value may fluctuate in response to general economic and market conditions and the perception of individual issuers.

Investments in equity securities are generally more volatile than other types of securities. An investment that is concentrated in a specific sector or industry increases its vulnerability to any single economic, political or regulatory development affecting that sector or industry. This may result in greater price volatility.

Risks associated with the Technology sector include increased competition from domestic and international companies, unexpected changes in demand, regulatory actions, technical problems with key products, and the departure of key members of management. Technology and Internet-related stocks, especially smaller, less-seasoned companies, tend to be more volatile than the overall market.

There are no guarantees that growth or value stocks will increase in value or that their intrinsic values will eventually be recognized by the overall market. The return and principal value of stocks fluctuate with changes in market conditions. The growth and value type of investing tends to shift in and out of favor.

An index is unmanaged and unavailable for direct investment.

S&P 500 Index consists of 500 stocks chosen for market size, liquidity, and industry group representation. It is a market-value-weighted index with each stock’s weight in the index proportionate to its market value.

Stanford University AI Index tracks, collates, distills, and visualizes data relating to artificial intelligence.

Global Securities Research (GSR) and Global Investment Strategy (GIS) are divisions of Wells Fargo Investment Institute, Inc. (WFII). WFII is a registered investment adviser and wholly owned subsidiary of Wells Fargo Bank, N.A., a bank affiliate of Wells Fargo & Company.

The information in this report was prepared by Global Securities Research (GSR). Opinions represent GSR’s opinion as of the date of this report and are for general information purposes only and are not intended to predict or guarantee the future performance of any individual security, market sector, or the markets generally. GSR does not undertake to advise you of any change in its opinions or the information contained in this report. Wells Fargo & Company affiliates may issue reports or have opinions that are inconsistent with, and reach different conclusions from, this report. Past performance is no guarantee of future results.

The information contained herein constitutes general information and is not directed to, designed for, or individually tailored to, any particular investor or potential investor. This report is not intended to be a client-specific suitability or best interest analysis or recommendation, an offer to participate in any investment, or a recommendation to buy, hold, or sell securities. Do not use this report as the sole basis for investment decisions. Do not select an asset class or investment product based on performance alone. Consider all relevant information, including your existing portfolio, investment objectives, risk tolerance, liquidity needs, and investment time horizon. The material contained herein has been prepared from sources and data we believe to be reliable, but we make no guarantee to its accuracy or completeness.

Global Securities Research works with information received from various resources including, but not limited to, research from affiliated and unaffiliated research correspondents as well as other sources. Global Securities Research does not assign ratings to or project target prices for any of the securities mentioned in this report.

Global Securities Research receives research from affiliated and unaffiliated correspondent research providers with which Wells Fargo Investment Institute has an agreement to obtain research reports. Each correspondent research report reflects the different assumptions, opinions, and the methods of the analysts who prepare them. Any opinions, prices, or estimates contained in this report are as of the date of this publication and are subject to change without notice.

Wells Fargo Advisors is registered with the U.S. Securities and Exchange Commission and the Financial Industry Regulatory Authority but is not licensed or registered with any financial services regulatory authority outside of the U.S. Non-U.S. residents who maintain U.S.-based financial services account(s) with Wells Fargo Advisors may not be afforded certain protections conferred by legislation and regulations in their country of residence in respect of any investments, investment transactions, or communications made with Wells Fargo Advisors.

Wells Fargo Advisors is a trade name used by Wells Fargo Clearing Services, LLC and Wells Fargo Advisors Financial Network, LLC, Members SIPC, separate registered broker-dealers and non-bank affiliates of Wells Fargo & Company.