Capital IdeasTM

Investment insights from Capital Group

Categories
人工知能
Separating AI hype from investment opportunity
Mark L. Casey
Equity Portfolio Manager
Peter Eliot
Equity Portfolio Manager
Jared Franz
Economist

The artificial intelligence hype cycle has kicked into hyperdrive. 


Newsfeeds bring daily reports of generative AI’s promise to accelerate medical and scientific discovery, drive leaps in productivity and eliminate jobs on a mass scale.


Businesses and individuals have rushed to adopt AI tools. OpenAI, developer of the ChatGPT chatbot, has reported that the tool had 100 million weekly users by the start of 2024, including two million developers and 92% of Fortune 500 companies.


Investor excitement has turbocharged share prices for the most visible enablers of generative AI, including NVIDIA, Meta Platforms and Microsoft, an investor in OpenAI. From 2023 through February 2024, the three tech giants soared 519%, 304% and 77%, respectively.


The question for investors: Has generative AI — platforms and tools that can generate new content including, text, images, video and sound — launched a lasting megatrend, or has its potential been greatly exaggerated?


We believe generative AI represents a transformational shift that will lead to unprecedented investment opportunities. The challenge for investors today is to separate the hype from what matters: the pace of adoption, improvements in the models and price declines.


The size of the market for AI is difficult to measure

This bar chart is labeled, “Annual installation of new industrial robots (‘000 units)” and visualizes the number of new industrial robot installations from 2017 to 2026. It distinguishes between traditional robots and collaborative robots from 2017 to 2022. The y-axis represents the number of installations in thousands, while the x-axis represents the years. The number of collaborative and traditional robots from 2017 to 2022 are as follows: In 2017, there were 11k collaborative robots and 389k traditional robots; in 2018 there were 19k collaborative and 405k traditional; in 2019 there were 21k collaborative and 366k traditional; in 2020 there were 26k collaborative and 363k traditional; in 2021 there were 42k collaborative and 484k traditional; and in 2022 there were 55k collaborative and 498k traditional. There is a caption that shows that there was a 31% growth in collaborative robot installations between 2021 and 2022. Additionally, there are forecasts for the total number of industrial robots from 2023 to 2026. The forecast is for 593k robots in 2023, 622k in 2024, 662k in 2025, and 718k in 2026.

Sources: Gartner, Statista, United Nations, World Bank. Smartphone market size estimate: total global population ages 15-64 (United Nations) multiplied by estimated average smartphone selling price of $291 (as of 2022). Digital payments market size estimate: total global GDP (World Bank) multiplied by estimated average take rate of 2% (as of 2022). Electric vehicles market size estimate: global automotive industry revenue (as of July 2023). Cloud/SaaS market size estimate: worldwide IT (information technology) spending (as of January 2023). SaaS: software as a service.

The AI market may be bigger than you think


With AI still in its infancy, it is difficult to estimate how large the market will be. The consulting firm McKinsey estimates that the AI market could reach $13 trillion by 2030, while PricewaterhouseCoopers assesses it could be $15.7 trillion by 2030. Meanwhile, Goldman Sachs has estimated it could boost GDP by 7%.


We believe the ultimate addressable market for AI is the hardest of any we have tried to size. After all, what is the “value” of better intelligence — or having more knowledge about everything? Simply put, AI has open-ended potential to generate a market many multiples larger than those of previous technological advances.


When the smartphone was popularised 15 years ago, combining internet and mobile phone technology, few could have predicted it would lead to new app-based business models like Uber and DoorDash. The breakthroughs in artificial intelligence represent similar potential.


What’s more, GenAI can analyse vast amounts of information and use pattern recognition to teach itself new tasks. This unlocks the potential to solve complex problems, uncover new insights and potentially lead to the creation of new business models.


Four ways early adopters are using AI


Companies are still in the experimentation phase with AI. Some IT consulting businesses, such as Accenture, have launched an initiative to help clients identify use cases for generative AI, adopt AI tools and integrate them into organisations.


Businesses are already seeking to tap into the efficiency and productivity gains AI can deliver. As much as 30% of hours worked in the US could be automated by 2030, according to the McKinsey study.


Here are just a few ways companies are integrating AI into their businesses.


1. Scientific and medical discovery


AI will likely accelerate discovery across various disciplines. Artificial intelligence can potentially process the totality of all know mathematics or physics or medicine in a very short time. For example, biopharma company Regeneron is harnessing GenAI tools to scour its massive database at its genetics center to identify disease targets, understand disease progression, develop drug therapies and track how individuals respond to treatment.


2. Product development


Consumer products giant Procter & Gamble has begun using GenAI to improve molecular discovery, enabling it to develop, for example, 100 fragrance options at a time rather than one. As a result, they can respond to new product trends within months instead of years.


Among consumer tech offerings, Meta has developed AI sunglasses equipped with cameras that see what the wearer sees and microphones to hear what the wearer hears. The glasses are designed to absorb information around the user, including capturing photos or making suggestions for things like spreadsheet formatting, for example.


3. Industrial automation and robotics


Amazon, which has long used industrial robots in its fulfillment centers, is applying AI to improve their efficiency and functionality. What’s more, robotics company Figure has harnessed ChatGPT to create a “humanoid” robot with audio and video inputs that can converse with and work alongside humans. Schneider Electric is developing a generative AI tool to communicate with customers about their carbon emissions. Construction and mining equipment manufacturer Caterpillar is investing in AI to boost productivity of its autonomous machines.


4. IT services


Companies were already shifting several of their IT needs to the cloud and outsourcing more of their tech support. This trend is being amplified by the proliferation of GenAI, which has spawned a host of smaller companies that offer larger enterprises consulting and IT services that can evaluate and score the quality of content being created through the AI process, which can generate false or misleading information, known as hallucinations.


Other companies have been using GenAI to support call centers, improve financial models, create marketing and advertising content, analyse legal contracts, and develop employee training materials.


AI investment opportunities


While GenAI adoption has been most aggressive in the tech and media industries, its use has spread across industries and organisations.


Businesses worldwide invested an estimated $19.4 billion in 2023 to integrate AI into their processes, according to a Wall Street Journal report. While many are in the experimentation phase, AI has the potential to create massive productivity gains, reduce costs for companies and generate insights that provide a competitive edge for early adopters.


AI can boost automation of a range of business tasks

The chart shows the percentage of survey respondents across a range of industries who have reported they are regular users of generative artificial intelligence, have used it once or have had no exposure to the technology. Results are as follows: within the technology, media and telecom industries, 50% regularly use, 37% tried at least once and 12% had no exposure; in financial services, 42% regularly use, 41% tried at least once and 18% had no exposure; in business, legal and professional services, 36% regularly use, 41% tried at least once and 23% had no exposure; in health care, pharma and medical products, 33% regularly use, 44% tried at least once and 22% had no exposure; within advanced industries, 32% regularly use, 47% used at least once and 20% had no exposure; within consumer goods and retail, 30% use regularly, 40% used at least once and 30% had no exposure; and within energy and materials, 29% use regularly, 50% used at least once and 22% had no exposure.

Sources: Capital Group, McKinsey Global Survey on AI. Figures may not sum to 100% due to rounding. Advanced industries include automotive and assembly, aerospace and defense and advanced electronics industries. Survey was conducted in April 2023 and included responses from 1,684 participants.

Wider adoption can fuel continuous improvements that build upon each other, setting in motion a virtuous cycle of innovation and adoption. As more enterprises adopt GenAI, new and more sophisticated uses will likely emerge.


Rather than focus on attention-grabbing headlines about massive job loss or concerns about the coming of sentient robots, investors should focus on declining costs of AI adoption, advancements in models and identifying early adopters that may use the technology to gain a competitive advantage.


As active investors, that’s where we’re spending our time to uncover investment opportunities that might match the vast potential of generative AI.



Mark L. Casey is an equity portfolio manager with 23 years of investment industry experience (as of 12/31/2023). He holds an MBA from Harvard and a bachelor’s degree in history from Yale.

Peter Eliot is an equity portfolio manager with 23 years of experience. He holds an MBA and a master's in international affairs from Columbia as well as a bachelor's in international relations from the University of Pennsylvania.

Jared Franz is an economist with 18 years of investment industry experience (as of 12/31/2023). He holds a PhD in economics from the University of Illinois at Chicago and a bachelor’s degree in mathematics from Northwestern University.


過去の実績は将来の成果を保証するものではありません。投資の価値および投資収益は減少することも増加することもあり、当初投資額の一部または全部を失うことがあります。本情報は投資、税務もしくはその他の助言の提供、または証券の売買の勧誘を意図するものではありません。

個人に帰属する記述は、その個人の出版日現在の意見を述べたものであり、必ずしもキャピタル・グループまたはその関連会社の意見を反映したものではありません。特に明記がない限り、すべての情報は記載された日付現在のものです。一部の情報は第三者から取得したものであり、そのため、かかる情報の信頼性については保証いたしません。