𝗔𝗜: 𝗧𝗜𝗡𝗔, 𝗙𝗢𝗠𝗢 𝗼𝗿 𝗙𝗘𝗠𝗢? 𝗝𝘂𝗻𝗲 𝟰, 𝟮𝟬𝟮𝟲 Artificial Intelligence remains one of the most powerful themes in today's financial markets. Despite the remarkable rise of many technology stocks, one question keeps coming back: are we witnessing a bubble, or the beginning of a sustainable growth cycle? To answer this question, it helps to distinguish between three forces currently driving AI-related investments. 𝟭 -- 𝗧𝗜𝗡𝗔: 𝗧𝗵𝗲𝗿𝗲 𝗜𝘀 𝗡𝗼 𝗔𝗹𝘁𝗲𝗿𝗻𝗮𝘁𝗶𝘃𝗲 In an environment marked by geopolitical uncertainty, concerns about global growth, and questions surrounding monetary policy, many investors are turning to AI because few alternatives offer comparable visibility. Companies linked to semiconductors, data centers, and digital infrastructure often benefit from: • Order backlogs extending several years • Strong visibility on future revenues • High profit margins In this context, it is only natural that capital continues to flow toward these businesses. However, if the global economic outlook improves and other sectors become more attractive again, part of these investment flows could shift toward currently overlooked opportunities. 𝟮 -- 𝗙𝗢𝗠𝗢: 𝗙𝗲𝗮𝗿 𝗢𝗳 𝗠𝗶𝘀𝘀𝗶𝗻𝗴 𝗢𝘂𝘁 The fear of missing the next technological revolution is very real. But unlike some past market bubbles, today's excesses do not appear to be driven primarily by retail investors. The real FOMO is taking place among hyperscalers and cloud giants. For them, AI is no longer simply an opportunity; it has become a matter of competitiveness. Failing to invest today could mean losing market leadership tomorrow. As a result, we are seeing: • Massive investments in data centers • Accelerated purchases of GPUs • Development of proprietary AI models • Construction of dedicated energy infrastructure Even if some projects eventually generate lower returns than expected, many of these companies believe that the risk of underinvesting is even greater. 𝟯 -- 𝗙𝗘𝗠𝗢: 𝗙𝗮𝗯𝘂𝗹𝗼𝘂𝘀 𝗘𝗮𝗿𝗻𝗶𝗻𝗴𝘀 𝗠𝗼𝗺𝗲𝗻𝘁𝘂𝗺 This is probably the strongest argument supporting the sector today. Many AI-related companies continue to report results above expectations: • Rapid revenue growth • Expanding profit margins • Rising earnings • Guidance upgrades quarter after quarter In other words, a significant part of the stock market appreciation is supported by genuine profit growth. That is what separates pure speculation from a trend backed by strong fundamentals. 𝗧𝗼𝗸𝗲𝗻𝘀: 𝗧𝗵𝗲 𝗞𝗲𝘆 𝗠𝗲𝘁𝗿𝗶𝗰 In my view, the most important data point today is neither stock prices nor the number of AI announcements. It is token consumption. Tokens represent the basic unit used by AI models. The more tokens are consumed, the more it indicates that: • Companies are actively using AI • Infrastructure is being utilized • Investments are generating real economic value Conversely, a sustained slowdown in token growth could signal that installed capacity is becoming excessive relative to actual demand. 𝗪𝗵𝘆 𝗜𝘀 𝗧𝗵𝗶𝘀 𝗦𝗼 𝗜𝗺𝗽𝗼𝗿𝘁𝗮𝗻𝘁? The entire AI value chain depends on continued growth in token consumption: • Lithography equipment manufacturers • Foundries • Memory producers • Chip designers • Cloud providers Today, some segments, particularly HBM memory used in AI systems, remain constrained by supply. This supports both pricing power and profit margins. However, over the long term, only growing AI adoption can justify the massive investments currently being made. For now, the momentum remains favorable. Profits continue to grow, investments keep increasing, and demand remains strong. However, one underestimated risk is the cost of computation. If the cost of using AI rises faster than the productivity gains it delivers, companies may begin to optimize and limit their usage. This is why I continue to closely monitor real token consumption. The key question is not how much companies are investing in AI, but how much AI is actually being used. $NSDQ100 $SPX500 $MU (Micron Technology, Inc.) ⚠️ Risk Warning: 𝘛𝘩𝘪𝘴 𝘪𝘴 𝘢 𝘱𝘦𝘳𝘴𝘰𝘯𝘢𝘭 𝘪𝘯𝘷𝘦𝘴𝘵𝘮𝘦𝘯𝘵 𝘷𝘪𝘦𝘸, 𝘯𝘰𝘵 𝘧𝘪𝘯𝘢𝘯𝘤𝘪𝘢𝘭 𝘢𝘥𝘷𝘪𝘤𝘦. 𝘞𝘩𝘦𝘵𝘩𝘦𝘳 𝘺𝘰𝘶 𝘤𝘩𝘰𝘰𝘴𝘦 𝘵𝘰 𝘢𝘤𝘵 𝘰𝘯 𝘪𝘵 𝘰𝘳 𝘯𝘰𝘵 𝘪𝘴 𝘦𝘯𝘵𝘳𝘦𝘭𝘺 𝘶𝘱 𝘵𝘰 𝘺𝘰𝘶. 𝘗𝘢𝘴𝘵 𝘱𝘦𝘳𝘧𝘰𝘳𝘮𝘢𝘯𝘤𝘦 𝘥𝘰𝘦𝘴 𝘯𝘰𝘵 𝘨𝘶𝘢𝘳𝘢𝘯𝘵𝘦𝘦 𝘧𝘶𝘵𝘶𝘳𝘦 𝘳𝘦𝘴𝘶𝘭𝘵𝘴.