Artificial Intelligence: Behind the Scene – Ecosystem

Before booking a table at a restaurant, you dive into reviews, learn about the chef, and explore the menu. Similarly, in AI, focusing solely on applications misses the bigger picture. The real story involves data, computing power, infrastructure, model researchers, and ethical guidelines. Understanding the whole thing is key to unlock the true potential and investment value of AI innovations.

Today, people and investors, the focus is almost entirely on applications like ChatGPT and OpenAI. We marvel at what AI can do—write essays, answer questions, or even hold conversations. But have you ever stopped to think about what makes these AI wonders tick behind the scenes? It’s not all magic; it’s an intricate ecosystem of infrastructure that keeps this digital powerhouse running. In fact, the real story of AI is much more than sleek applications and cutting-edge algorithms. Behind the curtain is a robust framework of hardware, software, and services, from data centres to cloud computing, batteries, and chips. And the tech world isn’t just about the Magnificent 7 (Meta, Apple, Amazon, Netflix, Alphabet, Microsoft, and Nvidia). Let’s dive deeper into the AI infrastructure that makes all of this possible and introduce some key players you might want to keep an eye on, including their stock performances.

The Heart of AI: Data Centres

Data centres are experiencing unprecedented demand due to the surge in AI workloads. According to a 2023 report from Synergy Research, the global data centres market is expected to grow by 10% annually over the next five years. The need for data centres is increasing due to the growth in AI applications, cloud services, and the Internet of Things (IoT). As AI models like ChatGPT become more complex, the servers required to run these models must expand in both size and capacity.

To meet this demand, companies like Equinix and Digital Realty Trust are rapidly scaling up their infrastructure. Equinix announced the construction of 29 new data centers in 2023 alone, and this expansion trend is mirrored across the industry. The global hyperscale data center count crossed 700 by mid-2023, with another 300 projected to be built by 2026.

Investing in data center REITs (Real Estate Investment Trusts), such as Equinix and Digital Realty Trust, can be a stable and growing market as AI continues to drive demand for more server space and greater capacity. Data centers aren't just real estate; they’re becoming the "factories" of the AI era.

The Fuel: Batteries and Energy Efficiency

AI workloads consume an enormous amount of electricity. A single data centres, running AI models like OpenAI’s GPT-4, can use as much electricity as 50,000 homes. In fact, the global electricity consumption for data centres in 2021 accounted for around 1% of the world’s total electricity use, according to the International Energy Agency (IEA). With AI workloads increasing, this figure is expected to grow substantially.

This immense energy demand highlights the crucial role of innovative battery and energy solutions. Data centers increasingly rely on advanced battery storage systems to ensure that power supplies remain stable, efficient, and environmentally sustainable. Companies like Tesla are developing megapack battery solutions that store excess energy and keep data centers running even during power shortages. Tesla’s energy division generated $6 billion in revenue in 2023, a sign that battery technologies are becoming critical for supporting energy-hungry AI infrastructure. While Tesla’s stock is more commonly associated with EVs, its energy division represents a growing opportunity, especially as data centre look to reduce costs and increase sustainability.

Vistra Corp (VST), one of the largest competitive power generators in the U.S., plays an increasingly vital role. Vistra has been investing heavily in both battery storage and nuclear energy as part of its strategy to meet growing electricity demand sustainably. Further, investors attach a higher value to the company’s nuclear power capacity to meet the growth in data centres.

‍

The Cloud: Hosting AI's Brainpower

The cloud is another cornerstone of AI infrastructure, and demand for cloud services continues to soar. Global spending on public cloud services is expected to reach $1 trillion by 2026, driven in large part by AI and machine learning workloads. Companies like Amazon Web Services (AWS), Microsoft Azure, and Google Cloud are investing billions to expand their AI capabilities. AWS alone is generated $90.8 billion in revenue in 2023, with a significant portion coming from cloud-based AI services. These cloud platforms provide the flexibility and scale required for AI applications to run efficiently. In fact, many startups and even large enterprises no longer buy servers, they simply lease space in the cloud, which can handle the spikes in demand that come with AI processing, they can use whenever they need, which lower the operating costs.

Cloud providers also offer various machine learning services, allowing companies to train their models without the need for in-house AI specialists. When you invest in the stocks of Amazon (AMZN), Microsoft (MSFT), or Google (GOOGL), you're not just betting on e-commerce, software, or search engines, you’re also riding the wave of AI-driven cloud infrastructure.

The Brains: AI Chips and Hardware

AI chips are the "brains" behind AI, and they are in massive demand. Nvidia, the leading provider of GPUs for AI, reported a revenue surge of 100% year-over-year in 2023, largely driven by sales of its AI chips. The company’s A100 and H100 GPUs are the gold standard, with data centres purchasing them in large quantities to power AI applications. Virtually, demand for Nvidia’s AI chips is so high that lead times for the latest H100 GPU are now over six months, according to industry reports.

Other companies like AMD and Intel are also seeing increased demand for their AI-focused hardware. AMD’s EPYC chips have begun to capture market share in cloud data centers, and Intel is heavily investing in developing AI-specific chips like the Gaudi2 to stay competitive.

The Soft Infrastructure: Software Tools

While hardware steals the spotlight, software tools are equally crucial to the AI ecosystem. Companies like Palantir and C3.ai provide the platforms that allow businesses to harness AI for real-world applications. Palantir, for example, reported a 21% increase in revenue in 2023, with its AI platform Foundry driving much of that growth. Palantir’s software is used across industries, from healthcare to logistics, enabling companies to make sense of complex datasets and improve efficiency.

C3.ai, another player in the AI software space, saw its revenue increase by 25% in 2023, driven by growing demand for enterprise AI applications. Its platform allows businesses to deploy AI models quickly, providing solutions for predictive maintenance, fraud detection, and supply chain optimization.

Beyond the Mag 7: Expanding the Tech World

While the "Magnificent 7" often dominate headlines, the AI infrastructure story goes beyond Meta, Apple, Amazon, Netflix, Alphabet, Microsoft, and Nvidia. Companies like Equinix (EQIX), NextEra Energy (NEE), Palantir Technologies (PLTR), and C3.ai (AI) represent the broader ecosystem that powers AI.

When we talk about AI's future, we shouldn’t just focus on the applications or the glamorous side of the technology. The real growth opportunity lies in the infrastructure that supports these advances. This isn’t just about what AI can do for us today but about how this underlying infrastructure will evolve to meet the demands of tomorrow.

So, the next time you ask ChatGPT a question, think about the intricate ballet happening behind the scenes, data centres humming, energy solutions powering through, clouds processing data, and specialized chips performing billions of calculations per second. This is the real engine of AI, and it's just as exciting as the applications we see on the surface.

‍