ChatGPT’s Dig at Musk’s Grok AI Ignites New OpenAI Code Debate

ChatGPT, OpenAI's popular AI chatbot, reignited controversy with Elon Musk this week by posting a screenshot suggesting Musk's AI model Grok gave a nearly identical response to a prompt. The incident renewed speculation that Grok was trained on OpenAI's code, though Musk denied this.

ChatGPT and Grok work similarly but differ in their data access. ChatGPT generates responses based on data up to 2021, while Grok, developed by Musk's company XAI, accesses some real-time data through social media platform X. The paid version of ChatGPT also provides real-time data.

Grok faced accusations of using OpenAI's code after refusing to provide information based on OpenAI's policy. XAI claimed this was unintentional, from scraping web data, and would fix in future Grok versions.

The AI race continues between companies like OpenAI and XAI, with disputes over how comparable their models are. ChatGPT's post suggested high similarity in a response from Grok that even mentioned OpenAI. Musk fired back, hinting ChatGPT may have scraped data from Grok for its own training.

XAI affiliated user Igor Babuschkin said the issue was rare and accidental, from scraping too much web data. "For what it's worth, the issue is very rare and now that we're aware of it we'll make sure that future versions of Grok don't have this problem," Babuschkin said. "Don't worry, no OpenAI code was used to make Grok."

The controversies highlight the complexities of developing and training AI systems, even for leaders in the field, as models become more advanced and data sources more vast and uncontrollable. With AI poised to impact our lives in so many ways, accountability and transparency will only become more important.