Juan Pedro Marín Arrese | A humble start-up company has dispelled the myth that AI development requires a substantial upfront investment. Spending only $5 million, it has achieved a better product than the leading semiconductor giants. Their shares sharply plummeted as investors realised that pouring so much money into their business proved too risky. A thorough reassessment of the expected returns has wreaked havoc on NVIDIA and other mega firms.
While no one questions AI will enjoy a bright future, the key issue raised by DeepSeek is the overall level of resources such an innovation might involve. Should they shrink compared to the vast amounts gauged up to now, a dramatic switch in costs, investments and returns would ensue. Most front runners might suffer substantial losses if their exposure grossly exceeds actual needs.
The disruptive impact of the newcomer will not lead to the collapse witnessed by the dot.com bubble at the beginning of the century. Yet, unless the main actors react swiftly, they are prone to experience a severe loss of confidence among investors.
At the end of day, what really matters is the economic impact. Should the breakthrough achieved by DeepSeek allow a substantial cut in the costs of AI delivery, both consumers and enterprises across the board would benefit. It will also help to enhance efficiency, increase competition and lower prices. Excellent news for users and better prospects of increasing both demand and offer for this new product.
Developing new products always involves an investment cost. As the alternative put forward by the newcomer lowers that bill, the economy as a whole will take advantage of such a move. Contrary to the widespread belief that investment adds value, it always involves a cost. Thus, only by trimming it to match the minimum necessary amount, does a real benefit emerge. Otherwise, the economy doesn’t get value for money.