The Bubble of the AI

November 16, 1999 Terra NetworksTelefónica’s Internet subsidiary, went over 13 euros surrounded by an unparalleled optimism. In a day he multiplied its price by three and allowed a few (and controversial) fortunate Millionaires instantly. Three months later he marked his historical mart, 157.65 euros. And then, the disaster.

The bubble of the Puntocom.

Terra was the Spanish paradigm of that phenomenon that shook the bags and especially the pockets of millions of people. In June 2003 Telefónica He ended up buying 100% Terra and paid 5.25 euros per share of investors, an almost symbolic figure taking into account what the company came to assert.

The company never had benefits in those years, and in 2002 it lost more than 2,000 million euros, four times more of its income. Terra was the Spanish Nvidia. At the peak its capitalization became more than 38,000 million euros: more than Repsol or the BBVA, and much more than the 24,000 that all telephone is worth today. However, Terra’s revenues in 2002 did not exceed 300 million euros. Does the story sound to you?

Probably yes. This is what some claim that it is currently happening with the segment of artificial intelligence, which from the appearance of Chatgpt has generated huge expectations. Even greater than the Internet in the late 1990s.

That has inevitably made You talk about an AI bubble. There are apparent signals everywhere:

1. Extraordinary assessments, tiny income: especially in the case of AI startupsthe valuations that companies like OpenAi have are colossal, but Your income is residual If we compare them with their losses. They are literally burning money as if there would be tomorrow.
2. Too much competition: The strategy of all these companies seems to go through Subscription models With plans with various prices, but the problem is hypercompetitiveness: There is no model that is clearly better than the restso here there can be a winner and many losers.
3. Expectations and realities. The managers of these companies are creating huge expectations On the impact of AI on our world, but tangible results are scarce.
4. Everything has to have ia. We are suffering the traditional entertainment of the term “AI”, which makes this technology lose part of its meaning: everything has to have if it is not the fundamental part of the product.
5. Regulatory and ethical challenges. And all this vertiginous development also faces the regulatory efforts that are especially strong in the European Union. Potential bad uses of technology (Deepfakes) or its negative consequences (Impact on employment) They are barriers for their development.

We also saw all that with the bubble of the Puntocom, and there are certainly more worrying indications that something weird is happening. The behavior of investors seems much more speculative than rational, media coverage – which tells us – shoots, or excessive trust in the future, despising potential risks, are clear. Is the bubble of AI a reality?

From the Puntocom to the Bubble Potential of the AI

There are differences between what happened then and what seems to be happening now. To begin with, AI seems to handle an equally solid or even more product than the Internet was when that bubble exploded: the impact is already being noticed in industries such as programming, for example.

In all cases the reality of AI is usually overwhelmed: this does not replace human work, But it enhances it. It is the new bicycle for the mindand those who use it – be in creative spheres or not – know that AI helps to do things faster and even better.

But unlike what happened with the Puntocom, this impulse of AI is much more starring the large technological companies, which are clear that they do not want anyone to advance them for the right.

Internet startups like Yahoo! Or Google surprised the giants of that time, and although there are now startups with an extraordinary assessment (OpenAi), it is also true that they will have a lot of competition from very based companies such as Microsoft, Alphabet or Nvidia, which has benefited especially from this explosion of interest in AI.

The potential of AI is evident, and is so clear or more than the one in the Internet then. But many startups of that bubble They did not have a clear business model: It was almost enough to buy a .com domain and have a good sales speech of the idea to get immediate financing because nobody wanted to stay out of that market.

Internet ended up having an almost unprecedented impact on the history of mankind, and created technological giants that supported themselves in that revolution. As they explain In five days“Neither everything was smoke, nor is it all solid.”

AI is generating colossal and even disturbing expectations, but it is not clear that Let’s be facing another bubble like that of the Puntocom. Time will end up responding to this unknown.

Image | Anne Nygård

In Xataka | The Copilot+ PC promised a revolution. I have tried one and for now there are more promises than realities