The big AI news of the week was Chinese startup DeepSeek releasing their AI model which seemingly offers similar functionality to ChatGPT. The news has apparently shocked the tech community “popping the US Big Tech bubble” and sent the share prices of several big tech players rocketing downward. This is the story that seems to be getting the most attention—how a startup from China could upend so much momentum and domination of the generative AI ecosystem that had previously been mainly US-based tech companies. There are two other elements to this story that are emerging and equally interesting to follow.
The first is just how deeply disruptive the technology behind DeepSeek seems to be in terms of its energy consumption. As one writer for MIT Tech Review observes, DeepSeek “poses a threat to the narrative that more computing power is the only thing that’ll unlock AI breakthroughs.” Even as innovation happens in terms of the speed and capability of these tools, there may be similar innovation in terms of how efficiently they can be run. The second story is thick irony of OpenAi’s complaint that DeepSeek used their proprietary content without permission … which is of course how their own tool was trained, using copyrighted intellectual property too.
So what should we take away from this whole emerging AI vs AI saga? Perhaps nothing long term. But it is interesting just how many backstories you can see inside something everyone is talking about if you look for them.
