It's impossible to gauge the return on investment for technology companies by simply looking at this year's or next year's financial statements. We haven't yet entered the phase where artificial intelligence can truly deliver transformation and value creation. There will be winners and losers, and it seems the stakes will be high on both sides. Do the claims of a bubble forming in the initial phase of the transformation created by AI reflect the market's sensitivity to speculative rhetoric, or do they indicate a lack of optimism about the outcomes of this transformation?

The AI ​​rally has suffered two major setbacks so far. The first was the shock of China's Deepseek developing a language model that could compete with its US rivals using different learning methods, with much more limited hardware and at a lower cost. If more capable competitors, without the massive investments of the "Magnificent Seven," could achieve effective results, were the valuations supporting the Nasdaq underpinned?

Wall Street questioned the return on AI investments for months. The second shock came with the wave of investment and financing from technology companies providing data center infrastructure. As analysts began questioning how technology companies' cash flows would finance investment announcements amounting to trillions of dollars, the initial reaction was felt in the valuations of smaller companies that these firms announced would meet their data center needs. OpenAI's announcements over several weeks, including astronomical service acquisitions and investments in some infrastructure providers, fueled concerns among market commentators about an "overinvestment and low-return trap." Google's Gemini3 launch further heightened these concerns, bringing to the headlines the idea that profitability in language models would be limited through competition.

As the year draws to a close, a battle of two opposing viewpoints continues among market players regarding AI investment. Those who believe a bubble has formed (e.g., Jamie Dimon and Michael Burry from JPMorgan) argue that all classic bubble indicators—overinvestment, overvaluation, overpositioning, and leverage—are currently valid. They emphasize that the rapid borrowing by large technology companies to finance these investments increases the risks (a debt crisis on top of capital pricing) that would arise if the bubble bursts.

Opposing viewpoints, such as Blackrock's, argue that unlike the dot-com bubble, the main players are generating high profits and cash flow, and that investments will increase revenue and profits and pay for themselves. Google's CEO, far from believing in a bubble, argues that the potential return for AI actors is not fully understood by the market, and therefore current valuations are conservative. Some analysts believe that the integration of AI into physical processes through robots has not yet been factored into prices. As is often the case, both approaches offer consistent and valid arguments. But do these arguments have a direct impact on current valuations and, more importantly, on future financial results? The investment plans announced by industry players contain astronomical figures for this initial phase of AI. However, like any investment, the investor's focus is on future market conditions rather than the present. We are only in the early stages of the economic and social transformation process that AI will bring about. Currently, artificial intelligence can classify complex and voluminous data, establish relationships between them and draw conclusions (Palantir, IBM), write (language models), code (Github, etc.), replace humans in customer communication (Salesforce, Ada, Genesys), support psychological therapies, and interact effectively with humans (as an assistant or friend). At this stage, artificial intelligence is carving out areas where the amount of service produced is limited by the high cost of labor used. The layoffs of software developers in the US and the increasing difficulty for new graduates in this field to find jobs are the first signs of this.