• ChatGPT app seen as the ‘iPhone moment’ for the AI industry
• Major Swiss bank issues an ‘upside revision’ and increases its AI growth forecast by 40%
• It believes AI will be ‘the tech theme of the decade’

Here’s some remarkable news for the start of 2024: In an update to investors and other interested parties, investment bank UBS has conceded it made a mistake in its previous estimation of the value of artificial intelligence (AI) to the global economy and admits that its earlier calculations were “too conservative” and wide of the mark. 

Yup, for the first time in living memory the gnomes of Zurich have voluntarily copped to a cock-up and recalculated their figures upwards by a massive 40%. Previously, the UBS team had forecast the value of global AI investments to grow from $28bn in 2022 to $300bn in 2027. Now though, “with visibility improving into how much firms are going to spend”, UBS has raised its forecast by 40% and expects AI spending in 2027 to be worth $420bn, which represents a compound annual growth rate (CAGR) of 72%. 

Much of the growth will come from increasing investments in the infrastructure that supports AI processing, which UBS expects to grow from $25.8bn in 2022 to $195bn in 2027, “driven by emerging trends like GPU cloud and AI edge computing.” Most, but not all, of that infrastructure spending will be on what UBS calls “AI computing”: In particular, spending on AI chips will be greater than earlier anticipated because of higher ‘inferencing’ (i.e. the application of rules to analyse and draw conclusions via the increased use of generative AI) and applications such as copilots.

At the same time, the investment bank’s equity investment team sees spending on AI applications and models growing from $2.2bn in 2022 to $225bn in 2027. It adds that “while such figures may seem lofty, they are consistent with prior phases of the computing cycle, such as mainframe, PC, and smartphone shipments.”

“This will likely make AI one of the fastest-growing and largest segments within global tech and arguably the tech theme of the decade, as we don’t see similar growth profiles elsewhere in tech,” noted Sundeep Gantori, equity strategist (global tech) in the UBS global wealth management team, in the bank’s new report, TechGPT: Raising AI revenue forecast by 40%.  

Central to the breezy rationale is the following premise: “If the launch of the ChatGPT application is the iPhone moment for the AI industry, the recent rollouts of numerous applications like copilots (digital assistants that help in the writing and debuting of code) and features like Turbo and vision from OpenAI in 4Q23 mean the App Store moment for the AI industry has arrived, in our view.”

What does this mean for the global tech sector? The UBS team notes that this year, the ever-increasing spending on AI infrastructure, and in particular AI computing products, will make the semiconductor manufacturers and platform companies that produce graphics processing units (GPUs) and other chips that are used to train and infer large language models (LLMs) the biggest beneficiaries of growth. As a result, investments in AI computing will increase more than tenfold between 2022 and 2027, from $15.8bn to $165bn. That’s good news in particular for Nvidia, which reported record levels of growth in 2023 and is expected to be the main beneficiary of AI computing spending for some time to come – see Nvidia’s sales triple as it capitalises on its AI advantage.

The other main chip firms set to benefit are AMD and Intel, while Arm is also keen to position itself as a key AI player.

But there are potential challenges that could impact the otherwise speedy growth trajectory of the AI sector. The main dark cloud that might appear on the wide blue yonder of the AI horizon is the “growing regulatory rhetoric around AI”, but UBS dismisses the threat as being highly unlikely to materialise “as increased regulatory focus should provide more opportunity than threat in 2024. Our overall view is that clearer regulatory frameworks in the earlier stages of a new technology’s development are preferable to retroactive regulation when a technology is well established... we think excessive corrections in major AI beneficiaries – due to geopolitics or regulations – could present a buying opportunity as underlying demand trends for AI should continue to be solid in the foreseeable future.” Thus, “the investment case for AI and related companies will endure in 2024 – and is in fact set to strengthen.” 

In summary, says UBS, regulations are worth monitoring as a risk factor, including export controls, but it adds, “as we highlighted recently in our semiconductor thematic note, an excessive correction due to geopolitics or regulations could present a buying opportunity.”

In a “Thought for the Day”, from the office of the CIO, UBS applauds US technology companies for ending 2023 as “the shining stars of their domestic market” and points out that “while the benchmark S&P 500 Index ended last year with a 24.2% price return, the so-called ‘Magnificent Seven’ stocks posted gains of between 48% and 249%. Many of these companies are integral players in the development and dissemination of artificial intelligence.”

The ‘Magnificent Seven’ are Apple, Microsoft, Alphabet (Google), Amazon, Meta, Tesla and Nvidia.  

The UBS Group has been growing its own AI workforce in recent years, recruiting data scientists, architects, business analysts and others because “the future of banking is in the cognitive bank.”

To keep up to date with all of TelecomTV’s coverage of the AI sector, check out our dedicated AI news page.

- Martyn Warwick, Editor in Chief, TelecomTV