The Artificial Intelligence (AI) investment boom is predicated on the idea that the industry will grow and mature over time to the point where the leading companies will transition to profitability over time. The typical evolution for such startups is that at the start they are dependent on outside investment to fund their operations and growth. Eventually, their sales should increase to the level that they become self-funded as their revenue becomes sufficient to fund their investment needs. Then eventually the revenue is expected to exceed their investment needs which will allow the companies to generate profits that will accrue to the investors that funded the early growth. This path to profitability is required for the AI companies to justify the large investment dollars they have attracted, and it must come quickly enough for early investors to achieve high enough annual rates of return to reflect the risks underwritten.
AI is a powerful tool and the optimism is justifiable, however it requires an abundant amount of power and data storage for it to run. Data centers are buildings filled with computers that store information and help run AI systems. Because AI is growing so fast, companies are trying to build more data centers. But there’s a big problem: there aren’t enough workers with the right skills to build and run them.
Many different types of workers are needed, like construction workers, electricians, engineers, cybersecurity experts, and cooling system technicians. But the number of trained workers isn’t keeping up with how fast new data centers are being built.
Another issue is that data centers are becoming more advanced. They use a lot more power than before, and that amount is growing very quickly. This makes them harder to design and manage. To help with this, companies are using special software to create “digital twins,” which are virtual models of data centers.
They also use parts that are built in factories and then put together at the site. This helps save time and money.
Even with these improvements, building data centers is still difficult. The United States already has a shortage of skilled construction workers, and many projects are delayed because there aren’t enough people to do the work. On top of that, getting permits and access to power can take a long time, which slows things down even more.
After the data centers are built, companies still need workers to run them. But it’s hard to find people for these jobs too, especially in areas like cybersecurity and AI. Since AI is still new, not many people have experience in it yet.
Keeping workers is also a challenge. Many people leave these jobs because they can be stressful and require long hours. Even though companies are paying more, workers also want better work-life balance and chances to grow in their careers.
To fix these problems, companies are trying to train their current workers, offer better benefits, and work with schools to prepare future workers. They are also using more technology and automation to help reduce the need for workers.
In the end, as AI keeps growing, the biggest challenge might not be the technology itself but having enough resources to support it. Meanwhile, a high growth rate will be needed for investors to achieve their target rates of return. Should the growth rate slow due to these various challenges, the path to eventual profitability may be lengthened, which will reduce the annual rates of return to AI investors and make AI investments look less attractive than the current hype that the industry is selling to investors.