Over the weekend, the Chinese introduced a new AI model called DeepSeek, which reportedly matches the performance of OpenAI's o1 but costs 27 times less.
The significantly lower development cost was confirmed by Microsoft’s CEO during the Davos summit.
DeepSeek is far cheaper and competes at the level of top AIs developed by OpenAI, $GOOGL, $META, $AMZN, and others that have spent tens of billions of dollars on their AI initiatives.
The market's reaction is clear: companies pouring massive resources into development are seeing their stocks drop (costs outweigh results), along with semiconductor companies like $NVDA, as spending on their products is expected to decrease.
DeepSeek’s performance is solid, but there’s nothing groundbreaking about it. The project's cost is impressive, but the model itself is based on existing architecture. They successfully leveraged and combined what had already been developed.
While Western companies have invested billions of dollars in R&D, the Chinese simply observed the results and applied the best solutions.
However, this is essentially the past, as the primary focus of the entire AI market is now the development of a so-called super AI — AGI, whatever that may entail.
Yes, they’ve made progress — achieving what top companies have already done, but more affordably.
But this is unlikely to indicate they’ve gotten any closer to significant breakthroughs.
All current AI products will continue to evolve and improve, so creating something cheap now is impressive but unlikely to remain relevant six months from now.
Bottom line: Stocks are plunging due to the market’s sharp reaction.
Will there be a rebound in today’s session? Possibly (the market is overly red). Could the decline continue? Also possible if the market keeps reacting so sharply to what is, in reality, not a game-changing development for the future of these companies.