Elon Musk’s X has implemented default settings of user posts and interactions on social media platforms to help train its AI chatbot, Grok.
“Your Grok posts, interactions, inputs, and results will be used for training and fine-tuning,” the setup statement reads.
However, the scheme comes with an option for users to opt out of the process.
The default setting is to automatically capture content for Grok, but “all X users can control whether their public posts can be used to train Grok, our AI search assistant,” the company said in a post on X (formerly Twitter) on Friday. “The setting is available on the web platform and will roll out to mobile soon.”
Users can also prevent their content from being sent to Grok by making their account private, the post said.
Grok is an AI chatbot modeled after OpenAI’s ChatGPT that was launched last fall to premium X subscribers and is part of Musk-owned AI startup xAI.
Musk paid $44 billion to buy Twitter and take it private in 2022. The deal, the lucrativeest ever for a social media company, was born out of Musk’s dissatisfaction with Twitter’s content moderation and other policies at the time. In the nearly two years since, he has significantly cut staff and rolled out a variety of monetization plans, including a premium subscription tier.
While Musk remains optimistic about AI as a key part of his business empire, which also includes SpaceX, he also signed an open letter in 2023 warning of the risks of AI. Written by 1,800 signatories, including Apple co-founder Steve Wozniak and other leading scientists, the letter called for a six-month halt to development of any system more powerful than OpenAI’s ChatGPT-4. It called for a six-month halt to development of any system “more powerful” than GPT-4. Engineers from Amazon, DeepMind, Google, Meta and Microsoft also joined the coalition, though some later backed out.
