Every major technological revolution tends to arrive with a promise: to change the economy, transform productivity, and open a new phase of growth. It happened with the internet in the 1990s. It is happening now with artificial intelligence.
The comparison with the dot-com bubble is inevitable. In both cases, an emerging technology captured the imagination of investors, companies, governments, and consumers. In both cases, markets began pricing in a future of accelerated growth. And in both cases an uncomfortable question appeared: are we investing in a real transformation or in a narrative that still has not demonstrated its returns?
Artificial intelligence is not a passing fad. Its impact on productivity, defense, education, health, financial services, automation, and trade will be profound. But a real technology can also generate a financial bubble. History shows that both things can be true at the same time.
The lesson of the dot-com bubble
The dot-com bubble did not burst because the internet was irrelevant. It burst because markets valued too quickly companies that still lacked sustainable business models.
Many firms promised to dominate the new digital economy, but they did not generate enough revenue, they had no profitability, and they depended on future expectations. When capital stopped financing promises and started demanding results, the market corrected abruptly.
And yet the internet did transform the world. After the collapse, companies survived that had built infrastructure, real business models, and durable competitive advantages. Amazon, Google, and other platforms did not deny the bubble; they showed that future winners can also be born inside a bubble.
The current question is similar: will artificial intelligence follow the path of the internet as a productive revolution, or the path of the dot-com era as financial excess?
What makes the current artificial intelligence cycle different
The comparison with the year 2000 has limits. Unlike many dot-com companies, the main actors in the current cycle are not startups without revenue. They are companies with cash, profits, infrastructure, global clients, and dominant positions.
Microsoft, Alphabet, Amazon, Meta, Nvidia, and other major technology players do not depend only on promises. They already have profitable businesses. In addition, they are financing much of their spending on artificial intelligence with their own resources, not only with debt or speculative capital.
This point matters. The current investment wave in artificial intelligence is supported by firms with real financial capacity. The cycle is not built only on fragile companies, but on corporations that are already part of the economic and technological core of the United States.
But this does not remove the risk. It shifts it. The problem is not whether artificial intelligence exists or whether it will matter. The problem is whether the economic returns will arrive as quickly as markets are already pricing in.
The new center of risk: capital spending and data centers
Artificial intelligence requires expensive infrastructure: chips, data centers, energy, cooling, networks, talent, and large-scale models. This has turned capital spending into one of the main indicators for understanding the cycle.
Artificial intelligence is no longer only software. It is physical, energy, and financial infrastructure. The problem is that this investment must generate returns. Companies are building capacity before there is full clarity about monetization.
If business and consumer demand grows at the expected pace, the spending may be justified. If not, the market may begin questioning the profitability of the infrastructure.
The key question is whether we are looking at a long-term strategic investment or at a spending race in which everyone invests out of fear of being left behind.
The difference between a technological bubble and a financial bubble
A technological bubble does not mean the technology is false. It means the price of assets separates from real economic results.
Artificial intelligence can transform industries and, at the same time, generate excessive valuations in certain companies. It can increase productivity in the long term and, at the same time, produce stock-market corrections in the short term. It can be a real revolution and a partial bubble at the same time.
This distinction is fundamental. The issue is not deciding whether artificial intelligence is real or whether it is a bubble. The right question is more precise: what part of current value is backed by real profits, and what part depends on future expectations?
The role of Nvidia and semiconductors
No company symbolizes this cycle better than Nvidia. Its position in chips for artificial intelligence turned it into a central actor in the new digital infrastructure. Demand for GPUs, accelerators, and specialized hardware boosted its revenue, margins, and market valuation.
But this concentration also creates vulnerability. When one company, or a very small group of companies, concentrates a large share of market enthusiasm, any change in expectations can generate amplified effects.
The history of technology shows that selling picks and shovels during an investment frenzy can be extremely profitable. But it also shows that, when the expansion phase slows, infrastructure providers can face sharp revisions in demand, margins, and valuation.
The question is not whether Nvidia matters. It does. The question is whether the market's current price already discounts a future that is too perfect.
Artificial intelligence, productivity, and political time
One of the main arguments in favor of the current cycle is that artificial intelligence can generate productivity gains. If companies reduce costs, automate processes, improve decisions, and create new products, economic growth could accelerate.
But there is a problem of timing. Financial markets tend to move faster than the real economy. Stock markets price expectations in months; productivity takes years to materialize. Infrastructure is built first; benefits are captured later.
This gap between financial expectations and economic adoption is one of the biggest risks in the current cycle. If benefits take longer than expected, markets may correct even if the technology keeps advancing.
Artificial intelligence may be in a phase similar to the internet before maturity: an inevitable technology, but with an uneven distribution of winners and losers.
Geopolitical implications
Artificial intelligence is not only a business phenomenon. It is a central dimension of global power.
The United States retains an advantage because of its capital ecosystem, universities, technology companies, advanced semiconductors, and digital platforms. China is trying to reduce technological dependence, develop its own models, strengthen its chip industry, and apply artificial intelligence to manufacturing, security, logistics, and services.
Europe is trying to regulate without losing competitiveness. Latin America observes the process from a more vulnerable position: as a consumer of technology, a provider of data, a recipient of investment, and a region with relatively low capacity in advanced digital infrastructure.
The dispute will not be only over language models or applications. It will be over data centers, energy, chips, standards, talent, cloud infrastructure, intellectual property, and the ability to integrate artificial intelligence into productive sectors.
The geopolitical question is clear: who will capture the value of artificial intelligence - those who use it, those who regulate it, or those who control its infrastructure?
What does it mean for Latin America
Latin America faces both an opportunity and a risk.
The opportunity lies in using artificial intelligence for productivity, education, health, agriculture, public services, mining, logistics, energy, and state transparency. The region can leap stages if it adopts AI tools strategically.
The risk is being trapped as a simple user of foreign platforms. If Latin America does not develop its own capabilities, it will depend on infrastructure, models, cloud services, and standards designed outside the region. That will limit its technological sovereignty and reduce its ability to capture economic value.
The region does not need to compete directly with the United States or China in foundational models. But it can develop sectoral applications, local data, technical talent, smart regulation, and strategic alliances.
The regional question is whether Latin America will use artificial intelligence to transform its productive structure, or only to consume imported technology.
What does it imply for the BRICS
For the BRICS, artificial intelligence is a test of strategic coordination.
China and India have scale, talent, and digital ecosystems. Russia retains scientific, mathematical, and cybersecurity capabilities. Brazil can play an important role in agriculture, energy, climate data, and digital services. South Africa and new members can contribute regional positions, strategic resources, and emerging markets.
The bloc, however, faces a difficulty: there is still no integrated technological architecture. There are common interests, but also asymmetries, rivalries, and very different levels of digital development.
The challenge for the BRICS is not only to declare technological cooperation. It is to build concrete mechanisms: research centers, financing for digital infrastructure, payment interoperability, data standards, talent formation, and cooperation in mature semiconductors.
The question is whether the BRICS can turn demographic and economic weight into coordinated technological capacity.
Possible scenarios
1. Artificial intelligence consolidates a new productivity phase
In this scenario, massive infrastructure spending is justified by rapid business adoption. Companies manage to monetize AI, cut costs, expand margins, and create new markets. Current valuations hold up at least partially.
2. Selective correction, not systemic collapse
Some AI-linked firms keep their value, while others fall because they lack real revenue. The market distinguishes between critical infrastructure, profitable applications, and purely speculative projects.
3. Financial bubble with real technology
AI keeps advancing, but market prices correct because expectations were excessive. It is a scenario similar to the internet: the technology survives, but many investors lose money.
4. Geotechnological fragmentation
The United States, China, Europe, and other blocs develop separate technological ecosystems. AI becomes a tool of geopolitical competition, differentiated regulation, and strategic control of data and infrastructure.
Conclusion
Artificial intelligence should not be analyzed as a fad or as an automatic guarantee of growth. It is a transformative technology, but it is also a powerful financial narrative.
The comparison with the dot-com bubble serves to recall one lesson: a technological revolution can be real and, at the same time, overvalued in markets. The internet changed the world, but not every internet company survived. Artificial intelligence may follow the same pattern.
The core issue is not whether AI matters. It does. The real question is who will capture the value, when returns will arrive, and which actors will be exposed if expectations outrun reality.
Are we at the beginning of a new era of productivity or in a phase of financial euphoria? Will big tech firms be able to justify massive infrastructure spending? Is Latin America prepared to capture value or will it only import outside solutions? Will the BRICS manage to build a technological agenda of their own or remain dependent on platforms dominated by others?
The answer will depend not only on technology. It will depend on investment, regulation, energy, talent, infrastructure, and geopolitical strategy. That is where it will be decided whether artificial intelligence becomes a shared productive revolution or a new concentration of global economic power.