Elon Musk – CEO of SpaceX, Tesla and, as of last October, Twitter – recently made headlines when he spearheaded a request to halt AI development due to concerns. Despite the effort, the tech mogul seems to be pushing forward with his own plans for developing AI infrastructure.
Now, a report from Business Insider revealed that Musk seems to be moving forward with plans for an artificial intelligence (AI) project at Twitter. According to two anonymous sources familiar with the company, the Twitter CEO recently purchased nearly 10,000 graphics processing units (GPUs) to be used on the platform.
Typically GPUs work on large-scale AI models due to the massive computation power required from the technology. This follows a tweet from Musk on March 18, in which he said that the company will use AI to “detect & highlight manipulation of public opinion” on Twitter.
In the months ahead, we will use AI to detect & highlight manipulation of public opinion on this platform.
Let’s see what the psy ops cat drags in …
— Elon Musk (@elonmusk) March 18, 2023
The anonymous sources reportedly deemed the project to be in its early stages, but that the purchase of such a quantity of GPUs shows Musk is “committed” to it. One of the sources said the project works with a large language model (LLM), however, the sources also said the definite role of generative AI at Twitter is still not totally clear.
Related: Bloomberg reveals AI for financial data, community responds
Additionally, Twitter recently acquired new talent with an extensive background in the AI field. In March, engineers Igor Babuschkin and Manuel Kroiss joined Musk’s team after working with DeepMind, an AI research subsidiary of Alphabet – Google’s parent company.
These developments come only a few weeks after Musk signed an open letter along with thousands of other researchers in the tech space to temporarily halt the development of AI due to risk for humanity.
Musk is also on record in 2017 giving a warning to regulators at an event with the U.S. National Governors Association that AI research needs to be regulated “before it’s too late.”