As the AI industry expands, investors are turning their focus from software to the essential physical infrastructure, including power, fuel, and heavy machinery, underpinning the sector’s growth.
The AI trade is increasingly looking less like a software story and more like an industrial one.
Nvidia chief executive Jensen Huang has been arguing that artificial intelligence should be seen as a “five-layer cake”: energy, chips, computing infrastructure, models and applications. The point, as he has framed it at the World Economic Forum in Davos and elsewhere, is that the flashy app layer rests on a far larger physical base. In that reading, the real money is not only in the visible AI products, but in the pipes, power lines and hardware needed to keep them running.
That shift is already filtering into markets. Semiconductor stocks have enjoyed the first wave of enthusiasm, with the sector now making up a sizeable share of the S&P 500. But investors are beginning to look further down the stack, towards companies that can supply the electricity, transport the fuel and build the electrical backbone for AI data centres.
Caterpillar offered one signal of how broad that demand can be. The company beat expectations and lifted guidance after revenue rose 22% from a year earlier, helped by AI-related infrastructure spending. That suggests the artificial intelligence boom is not confined to coding labs and cloud platforms; it is also flowing into heavy equipment, construction and industrial supply chains.
Among the names attracting attention are Williams, Eaton and Cameco. Williams, which moves about 16% of US natural gas, has started work on a new pipeline, underscoring how power-hungry data centres are reshaping demand for fuel and transport infrastructure. Eaton, the electrical equipment group, has now posted 16 straight quarters of growth in hyperscaler bookings, with its order book stretching into 2027. Cameco, the largest pure-play Western uranium miner, has also benefited from the new appetite for reliable power after four big tech companies signed nuclear-related deals over the past 18 months.
The common thread is simple enough: AI may be sold through software and semiconductors, but it is being built on a much older foundation of steel, fuel and electrons. For investors, that means the most durable opportunities may sit not in the brightest part of the boom, but in the less glamorous machinery beneath it.
- https://readthejoe.com/markets/the-ai-boom-needs-pipelines-more-than-python/ – Please view link – unable to able to access data
- https://blogs.nvidia.com/blog/davos-wef-blackrock-ceo-larry-fink-jensen-huang/ – At the World Economic Forum in Davos, Jensen Huang, CEO of NVIDIA, described AI as the foundation of the ‘largest infrastructure buildout in human history,’ encompassing energy, chips, computing infrastructure, cloud data centers, AI models, and applications. He emphasized that AI is not just a single technology but a comprehensive ‘five-layer cake’ that requires development across all layers to drive economic growth and job creation. ([blogs.nvidia.com](https://blogs.nvidia.com/blog/davos-wef-blackrock-ceo-larry-fink-jensen-huang/?utm_source=openai))
- https://blogs.nvidia.com/blog/ai-5-layer-cake/ – Jensen Huang, CEO of NVIDIA, articulated a vision of AI as a ‘five-layer cake,’ consisting of energy, chips, infrastructure, models, and applications. He highlighted that AI is not merely a clever app or a single model but essential infrastructure, akin to electricity or the internet. This perspective underscores the need for substantial investment and development across all layers to fully realize AI’s potential. ([blogs.nvidia.com](https://blogs.nvidia.com/blog/ai-5-layer-cake/?utm_source=openai))
- https://www.cnbc.com/2026/04/30/caterpillar-earnings-ai-demand-boosts-revenue-22percent.html – Caterpillar Inc. reported earnings that exceeded expectations, raising its outlook as AI demand contributed to a 22% increase in revenue compared to the previous year. This growth underscores the significant impact of AI on various industries, driving demand for infrastructure and related services.
- https://www.marketwatch.com/2026/04/30/williams-etn-ccj-ai-infrastructure-investments.html – Three infrastructure companies—Williams, Eaton, and Cameco—are gaining attention due to their investments in AI-related infrastructure. Williams, which transports approximately 16% of U.S. natural gas, has initiated a new pipeline project. Eaton, a leading electrical equipment manufacturer, has experienced 16 consecutive quarters of growth in hyperscaler bookings, with backlogs extending into 2027. Cameco, the largest pure-play Western uranium miner, has secured nuclear deals with four major tech firms over the past 18 months. These developments highlight the expanding role of infrastructure companies in supporting the AI industry’s growth.
- https://www.aiarainia.com/en/news/industry/nvidia-ai-five-layer-cake – NVIDIA CEO Jensen Huang has articulated a vision of AI as a ‘five-layer cake,’ comprising energy, chips, infrastructure, models, and applications. He emphasized that AI is not merely a clever app or a single model but essential infrastructure, akin to electricity or the internet. This perspective underscores the need for substantial investment and development across all layers to fully realize AI’s potential. ([aiarainia.com](https://www.aiarainia.com/en/news/industry/nvidia-ai-five-layer-cake?utm_source=openai))
- https://www.techradar.com/pro/nvidia-wants-to-have-your-cake-and-eat-it-jensen-described-the-ai-layered-stack-and-hints-at-what-worlds-most-valuable-firm-will-do-next – NVIDIA CEO Jensen Huang has outlined a comprehensive framework describing artificial intelligence as a five-layered industrial ecosystem, consisting of energy, chips, infrastructure, models, and applications. Each layer supports the operations of those above it, from physical power plants to end-user AI applications like self-driving cars and robotics. NVIDIA already dominates the chip and infrastructure layers, operating expansive data centers—termed ‘AI factories’—that require significant investments in land, energy, and computing systems. Huang emphasized that trillions of dollars in infrastructure investment are still needed to fully realize global AI capabilities. ([techradar.com](https://www.techradar.com/pro/nvidia-wants-to-have-your-cake-and-eat-it-jensen-described-the-ai-layered-stack-and-hints-at-what-worlds-most-valuable-firm-will-do-next?utm_source=openai))
Noah Fact Check Pro
The draft above was created using the information available at the time the story first
emerged. We’ve since applied our fact-checking process to the final narrative, based on the criteria listed
below. The results are intended to help you assess the credibility of the piece and highlight any areas that may
warrant further investigation.
Freshness check
Score:
7
Notes:
The article discusses Nvidia CEO Jensen Huang’s ‘five-layer cake’ framework for AI infrastructure, first introduced at the World Economic Forum in Davos in January 2026. ([blogs.nvidia.com](https://blogs.nvidia.com/blog/davos-wef-blackrock-ceo-larry-fink-jensen-huang/?utm_source=openai)) The earliest known publication date of this concept is January 21, 2026. The article also mentions companies like Caterpillar, Williams, Eaton, and Cameco, which have been discussed in various contexts over the past few months. ([scmp.com](https://www.scmp.com/tech/tech-war/article/3340709/watch-nvidia-ceo-jensen-huang-speak-davos?utm_source=openai)) However, the specific combination of these elements in this narrative appears to be original. Given that the core concept was introduced in January 2026, and the article was published in May 2026, the freshness score is reduced due to the time elapsed.
Quotes check
Score:
6
Notes:
The article attributes a direct quote to Jensen Huang: ‘AI is a five-layer cake: energy, chips, computing infrastructure, models and applications.’ This quote was first reported in January 2026 during the World Economic Forum in Davos. ([blogs.nvidia.com](https://blogs.nvidia.com/blog/davos-wef-blackrock-ceo-larry-fink-jensen-huang/?utm_source=openai)) The article does not provide a direct source for this quote, making independent verification challenging. The lack of a direct source for the quote raises concerns about its authenticity.
Source reliability
Score:
5
Notes:
The article is published on ‘readthejoe.com,’ a niche publication with limited online presence and credibility. The lack of a direct source for the quote attributed to Jensen Huang further diminishes the reliability of the information presented. The absence of a reputable source for the quote raises concerns about the accuracy and authenticity of the information.
Plausibility check
Score:
7
Notes:
The article discusses Nvidia’s role in AI infrastructure and mentions companies like Caterpillar, Williams, Eaton, and Cameco, which are involved in sectors related to AI infrastructure. However, the specific combination of these elements in this narrative appears to be original. The plausibility of the claims is reasonable, but the lack of direct sources for key statements reduces confidence in their accuracy.
Overall assessment
Verdict (FAIL, OPEN, PASS): FAIL
Confidence (LOW, MEDIUM, HIGH): MEDIUM
Summary:
The article presents a narrative that combines Nvidia’s ‘five-layer cake’ framework for AI infrastructure with mentions of companies like Caterpillar, Williams, Eaton, and Cameco. However, the lack of a direct source for the quote attributed to Jensen Huang and the reliance on a niche publication with limited credibility raise significant concerns about the accuracy and authenticity of the information. The absence of independent verification sources further diminishes confidence in the content’s reliability. Given these issues, the article does not meet the necessary standards for publication.