_edited.png)
KAPITALKOMPASS #51: AI boom or bubble?
- service4100
- Oct 28
- 4 min read
Updated: 3 days ago
Dear readers,
Artificial intelligence is the investment theme of the year—a balance between substance and speculation. Many are asking: Is the pattern of the dot-com era repeating itself?
Our analysis identifies parallels, differences, and consequences for investing – concisely, fact-based, and with clear courses of action.
The parallels:
Euphoria, capital, exaggeration
At the end of the 1990s, the internet seemed to override all economic logic. Companies with a simple website received valuations in the billions – even without profit, often without even revenue. The belief in a new, "digital economy" drove investors and founders alike into a frenzy.
The pattern repeats itself: Since the success of generative AI models like ChatGPT and Claude, investors have been frantically searching for the next growth story. According to World Bank estimates, over $80 billion flowed into AI startups in 2024 alone—more than any other technology sector. Companies that include "AI" in their descriptions are sometimes experiencing dramatic share price gains.
As was the case back then, a new industry is flooded with money long before it has to prove that its products are actually profitable.
The differences:
Mature technology, real use
But unlike 25 years ago, today's boom is built on a more solid foundation.
The internet was a vision in 1999. By 2025, AI is already a tool. Applications have become a reality – in medicine, software design, financial analysis, and marketing. Millions of people use AI-based systems every day, and major corporations like Microsoft, Google, Amazon, and Nvidia are building entire business models around them.
The infrastructure exists: cloud data centers, specialized chips, massive amounts of data, and a digitally driven society. While many dot-com companies failed due to the limitations of technology, the AI industry today has the necessary foundation to at least partially fulfill its promises.
The risks:
Overvaluation and energy costs
Nevertheless, the AI boom is not without exaggerations. Some companies and startups are operating at high losses, with no clear path to profitability. Some lack a truly unique selling point—they rely on the infrastructure of the major players instead of developing their own technology.
Added to this is an economic factor that played hardly any role during the dot-com era: energy. Training and running large language models consumes enormous amounts of electricity. Studies estimate that a single AI model consumes as much energy per year as a medium-sized city. In the long term, rising energy costs and environmental regulations could become a tough selection criterion.
The real comparison:
Hype meets substance
After the dot-com bubble burst, a few powerful winners remained: Amazon, Google, and eBay. The rest disappeared. A similar scenario is conceivable in the AI sector. The industry will likely experience consolidation: several providers will disappear, and a few – supported by capital, data, and infrastructure – will dominate. For investors, this means that those who focus on short-term returns are taking a high risk. Those who focus on structural change in the long term are likely to profit.
Because AI is not just a product, but a basic technology, comparable to electricity or the internet. It will permeate almost every industry in the coming years—less spectacular, but all the more sustainable.
Conclusion:
No crash, but a correction
The dot-com bubble demonstrated that technological revolutions rarely follow a straightforward path. First comes euphoria, then disillusionment—and finally, integration into everyday life.
The AI hype will also go through these phases. Some companies will fail, capital flows will normalize, and the headlines will cool down. But the technology itself will endure—and with it those who have learned to use it economically.
The AI era may start out overhyped – but it will transform the economy and thus the productivity of companies.
An investment along the value chain of AI (chip designers, chip producers, data center operators, software developers as well as data security and power generation) seems sensible.
Here, too, diversification through ETFs is the right approach. Investing in individual companies is too risky.
Best regards and successful investing,
Your service team
Disclaimer
Important legal notice: The information contained in this newsletter is for general information purposes only and does not constitute investment advice or other professional advice. The data and analyses provided here are based on sources we consider reliable; however, we assume no liability for their timeliness, accuracy, completeness, or quality.
Investments in financial markets involve risks, including the possible loss of invested capital. Past performance is not an indicator of future results. Decisions based on the information contained in this newsletter are the sole responsibility of the reader. We assume no liability for any direct or indirect loss or damage that may arise from the use of this information. This newsletter should not be construed as an offer or solicitation to buy or sell securities or other financial instruments. We recommend seeking professional advice and considering the relevant legal and tax aspects before making any investment decision. The contents of this newsletter are protected by copyright. Any distribution, reproduction, or other use of the contents requires the prior written consent of the publisher.
Source:
Torsten Leissner