Despite his public criticism of artificial intelligence systems, Elon Musk, the owner of Twitter, has reportedly been working on his generative AI model project. 

According to sources, Musk has acquired thousands of GPU accelerators for the platform, indicating his serious intentions to implement his AI project. The AI project is reportedly still in its early stages, but sources suggest it includes a large language model (LLM) for Twitter, which has many training materials.

Musk’s acquisition of additional computing power suggests he is committed to developing the AI project. It is not yet known exactly what Twitter will use the generative AI for, but it is speculated that it could be used to improve search or advertising technologies. 

NVIDIA, which dominates the GPU accelerator market, offers options for large AI models that cost around $10,000 per instance. It is estimated that Musk has spent tens of millions of dollars to purchase the necessary equipment.

Elon Musk is one of the founders of OpenAI, a research institute dedicated to developing safe AI. However, he recently signed an open letter calling for a temporary moratorium on developing more complex AI models than GPT-4, citing the potentially disastrous consequences of uncontrolled AI development.

Musk’s involvement in AI development for Twitter is significant, given his previous criticism of AI systems. The acquisition of thousands of GPU accelerators and recruitment of AI development professionals suggests that Musk is serious about the project. However, given his concerns about the uncontrolled development of AI, it remains to be seen how AI will be implemented in the social network.