Musk was exposed to call for the suspension of AI research on the surface, secretly buying tens of thousands of GPUs to promote his own projects.
On April 12, according to two people familiar with the matter, although Tesla CEO Elon Musk (Elon Musk) signed an open letter calling for a six-month moratorium on advanced artificial intelligence research, he secretly Promoted the new generative artificial intelligence (AIGC) project at Twitter, and purchased tens of thousands of GPUs (graphics processing units) for this purpose .
Advertisement
According to foreign media citing people familiar with the matter, since the acquisition of Twitter was completed in October last year, Musk has made major reforms to the social media platform in many aspects. Recently, he bought about 10,000 graphics processors for Twitter. Given the computational workload required by new technologies, technology companies often use GPUs to process large AI model data.
Musk’s artificial intelligence project is still in its early stages, though the purchase of so much extra computing power suggests he has high hopes for the project, people familiar with the matter said. However, Musk has previously criticized recent developments in AIGC technology, saying the technology is so powerful that it needs regulation to ensure it operates in a “public interest . “
According to a person familiar with the matter, the artificial intelligence project that Musk is promoting on Twitter involves a large language model (LLM). LLM is still essentially an artificial intelligence that needs to be trained on large amounts of data in order to eventually create new and seemingly quite complex content and texts on its own. There is a huge amount of data on Twitter to train this kind of artificial intelligence model. For example, OpenAI, the creator of ChatGPT, previously had access to Twitter data for training purposes, although Musk said he blocked the practice in December.
Musk also brought in new artificial intelligence talent for Twitter . As of early March, he had formally hired engineers Igor Babuschkin and Manuel Kroiss from DeepMind, Alphabet’s artificial intelligence research subsidiary. Since at least February, Musk has been approaching people in the artificial intelligence field, including Babushkin, hoping to start his own artificial intelligence project to compete with ChatGPT.
It is unclear what Twitter will do with the AIGC, the people said. One use could be to improve search , which Musk has complained about as a poor experience. He even hired security developer George Hotz to advance a three-month project to “fix” the search function. However, Hotz left after only one month on the job.
People familiar with the matter also revealed that another possible use of Twitter AIGC is advertising , which is the biggest pillar of social media platform revenue. Twitter has struggled to court advertisers since Musk took over and changed many of its policies. After training, the AIGC has the ability to create new advertising images and text to target specific audiences.
GPUs can be very expensive these days. Nvidia, which has an estimated 95 percent market share, makes GPUs for large AI models for $10,000 (currently about 68,800 yuan). Musk has repeatedly said Twitter is on a shaky financial footing, but he likely spent tens of millions of dollars to add so many new GPUs, people familiar with the matter said .
The new project is expected to operate from one of Twitter’s two remaining data centers, most likely the one in Atlanta, the people added. Another data center is located in Oregon. Previously, Twitter’s main data center was located in Sacramento and was operated by NTT, but Musk finally decided to close it abruptly at the end of December last year. Behnam Rezaei, head of engineering at Twitter, also left.