Zuckerberg's $200 Billion Bet on AI

Meta had earmarked massive investment for developing its first advanced AI model, Muse Wonder, but CEO Mark Zuckerberg was not certain of the project's success, even with a potential cost of $200 billion.

Meta had considered a huge investment for the development of its first advanced artificial intelligence model, Muse Wonder; But Mark Zuckerberg, the CEO of the company, was not sure of the success of this project, even with the possible cost of 200 billion dollars.

According to Ambassador Arjan, quoted by Zomit, Zuckerberg talked about his doubts about the outcome of this investment in a conversation with Alexander Wang, the chief artificial intelligence manager of Meta. At the time, the project's success was by no means certain, Wang says.

Meta's internal tests show that this model is at the level of Fable 5.1 and Astra's GPT 6 in terms of performance. Also, in the MemoryBench test, it has recorded a better performance than Fable, Opus and Astra.

The cost of using this model is $4 per one million input tokens and $20 per one million output tokens; A number that puts it in the price range of Opus 5.5.

Mius Vander trained at the Prometheus Meta data center in Ohio. However, the results of independent evaluations of the model's performance have not yet been published, and claims about its capabilities cannot be independently verified.

Meta's huge investment shows that this company is willing to accept a big risk to compete in the field of advanced artificial intelligence; But determining the success rate of this bet depends on the publication of independent results and the actual performance of the model.