AI boom is fundamentally different from dot-com bubble

The bursting of the dot-com bubble in 2000 was a hugely important development. Some commentators and analysts are now attempting to frame the present artificial intelligence boom through that historical lens. Such comparisons, however, are inaccurate.

It has been argued that the enthusiasm surrounding AI resembles the speculative landscape that preceded the bursting of the dot-com bubble. But this view is detached from the economic reality of AI.

Some technological cycles and booms do contain speculative elements, but the current AI boom is not merely a repetition of the internet bubble.

Yes, the dot-com era was defined in large part by expectations and speculations. But by contrast, the rise of AI is occurring in a world where there is already an infrastructure for deploying it. Secondly, it is already generating tangible outcomes.

AI is not just a promise that may happen in the future. It is already reshaping sectors such as production, services, logistics, healthcare, finance, education and governance. As a result, this is not just a replay of 2000 but rather the early phase of a structurally transformative technological shift.

While the dominant perception of the late 1990s that the internet would change communication, media and other sectors was correct, the infrastructure to support such a large shift was not in place.

This is not just a replay of 2000 but rather the early phase of a structurally transformative technological shift

Dr. Majid Rafizadeh

For example, we did not have a digital system or smartphones or the cloud, while networks were underdeveloped. This means that while the internet was revolutionary, we did not have the infrastructure to translate its potential into reality. In other words, the markets had anticipated the future too quickly.

The current AI boom is emerging under profoundly different circumstances. Firstly, we already have high-speed connectivity, hyperscale cloud computing, advanced semiconductor manufacturing, sophisticated software ecosystems, mature digital payment systems and billions of digitally connected users. We do not have to wait to be technologically ready before using AI.

Secondly, we can examine the revenue profile of leading participants. The dot-com era was heavily populated by unprofitable startups that were dependent on speculative financing. But AI is led, to a significant extent, by some of the most profitable corporations in modern economic history. Firms such as Microsoft, Alphabet, Amazon, Meta, Nvidia and Oracle are financing large-scale AI investment from existing cash flows, diversified revenue streams and robust balance sheets. This shows how unlikely it is that these investments are just speculative.

Thirdly, AI is already producing measurable outcomes. Many organizations are using it to lower costs, increase efficiency and improve decision-making, as well as to expand service capacity.

For example, in healthcare, AI systems are helping with various issues, such as radiology analysis, medical imaging, clinical documentation and appointment scheduling. Hospitals and healthcare networks are increasingly adopting these tools to reduce administrative burdens and improve operational efficiency.

In finance, AI is being used for fraud detection, anti-money laundering reviews, credit scoring, compliance monitoring and customer service automation. Financial institutions are not investing in AI because it is trending but because they are seeing tangible results.

When it comes to manufacturing, Germany’s industrial sector, Japan’s robotics ecosystem, South Korea’s electronics manufacturers and China’s smart factories all show how AI has become an indispensable part of growth.

Even in education, AI tutoring systems, language learning assistants, grading tools and curriculum generators are changing how students are learning and how instruction is delivered.

AI has applications across nearly every sector. Some even argue that, in that respect, it is similar to electricity

Dr. Majid Rafizadeh

All these examples demonstrate the crucial issue that AI is not limited to one industry. It has applications across nearly every sector. Some even argue that, in that respect, it is similar to electricity.

While the internet certainly had transformative potential in the late 1990s, AI today can often be inserted directly into existing workflows. It enhances processes already taking place inside firms and institutions.

Fourthly, the dot-com bubble was strongly centered on US equity markets and startup culture, though its effects were broader. The current AI expansion is occurring across multiple regions.

While the US remains central, China is aggressively deploying AI across manufacturing, logistics, e-commerce and financial technology. Europe is focusing heavily on industrial AI and advanced manufacturing. India is leveraging AI within its vast information technology services sector. Japan and South Korea are integrating AI into robotics and automotive production. And Gulf states are investing in sovereign-scale AI infrastructure, data centers, smart city initiatives and diversification strategies. This worldwide competition is critical.

Some people might criticize the massive spending on data centers and chips, but it is vital to remember that, historically, transformative technologies often require large up-front investment.

Finally, labor markets also reveal that AI capability is already economically meaningful. Firms across many sectors are using AI to automate customer support; therefore, it is affecting real tasks now, not merely as a speculative narrative. In other words, AI has crossed an important threshold from demonstration to implementation.

In a nutshell, the current AI boom is not the same as the late-1990s dot-com bubble. AI is already generating tangible outcomes and is a tool that governments are increasingly funding, firms are deploying, scientists are using and industries are integrating.

• Dr. Majid Rafizadeh is a Harvard-educated Iranian-American political scientist. X: @Dr_Rafizadeh