The era of AI is already going very strong, and everyone knows that this brings big business behind it. The most popular, and the first one that we all probably tried, is ChatGPT. Well, its creator wants to continue evolving on the path of Artificial intelligence but without depending on anyone.
That is why today, thanks to a Reuter exclusive, we know that they would be interested in make your own chips of AI, and thus not depend on NVIDIA or any other brand, which, after all, are competition.
Will OpenAI manufacture your AI chips?
Apparently, the executives of this company would have been thinking about this option for some time, becoming self-sufficient and starting to manufacture their own components necessary for the AIs they own to advance in greater steps than the competition. But that is much more complicated if you depend on another company, in this case NVIDIA, who would be at the head of the design and production of these chips and which often fails to produce the demand it has.
The CEO of OpenIASam Altman was the first to report this chip shortage problem a long time ago, so we assume that he has been thinking about this idea ever since.
And NVIDIA controls 80% of this market, so diversifying and opting for another company can also be very complicated as well as risky.
Furthermore, ChatGPT does not require just any chip, but rather the most advanced ones, which translates into very high costs and greater production problems. Something that they could solve very easily by manufacturing them themselves, since using these products manufactured by NVIDIA would bring them closer to a cost of approximately 48.1 billion dollars in GPUs and around 16 billion dollars in chips per year to be able to supply all the queries of the users.
It is not clear whether they will finally carry out their plan or not, due to the fact that, obviously, the cost of building a factory to make this type of chips is not easy or cheap, hence there are not many companies and NVIDIA is is carrying the market. However, it may pay off in the long run. It will be something that they will have to continue meditating on, but if this is on the table, it is because there is a good chance it will happen.
The only quick and viable solution would be to acquire an already built chip factory, that would greatly improve the deadlines to start the project, as well as the costs of not having to start from scratch.
Who was going to tell the CEO of a small company that he was going to mess up so much with an application like ChatGPT. We’ll see how far all this goes. 2024 is positioned as the year of revolution in this field, so a lot of news like this awaits us. I hope they are all this interesting and we see things we never imagined. The world of AI has a lot of potential.
And you, what do you think of all this? Do you think this company will go it alone, trying to be the queen in this sector?
