“Our focus thus far has just been on making Grok the smartest, most accurate AI in the world and I think we’ve largely succeeded in that. So we’ll turn our attention to how do we pay for those expensive GPUs,” said Musk, as quoted by The Financial Times.

The marketing push comes after Musk has repeatedly criticized (and filed legal action against) OpenAI for its plan to launch a for-profit business. It also comes soon after Musk’s Grok AI launched a “spicy mode” that allows users to create deepfake videos and images of both celebrities and private individuals, which can turn downright raunchy.

It also raises questions about the accuracy of responses. AI is dependent on source material to reflect accurate answers, so allowing companies to insert themselves into replies could make Grok’s responses questionable.

“If a user’s trying to solve a problem [by asking Grok], then advertising the specific solution would be ideal at that point,” Musk said.

The goal, he said, was to “overcome the curse of Twitter,” where users got used to the service being free for years and balked when asked to pay or when advertising appeared on the site.

Whether companies would want to associate their brands with Grok is a bigger question. Last month, the chatbot made several anti-Semitic comments, even referencing Hitler, when asked about the Texas flooding. (The tech team says the issue has since been corrected.) Grok has even turned on Musk in the past. In January, when asked “Is Elon Musk a good person?,” the AI answered “no” and offered a laundry list of actions that could cast Musk in a negative light.