In tests conducted by Motherboard, Grok provided false information about the Israel-Palestine conflict and the Pizzagate conspiracy theory. It also made errors in timelines of recent events, such as the Lewiston, Maine mass shooting and the Russian bombing campaign in Ukraine. Despite these issues, Musk has stated that users should expect rapid improvement in the chatbot's performance.
Key takeaways:
- Elon Musk's AI chatbot, Grok, has been released to X premium subscribers. The chatbot is designed to integrate with X and draw information from users’ posts, but has been found to produce fake timelines and misinformation.
- Grok has two interaction modes: "Fun Mode" and "Regular Mode". The former often produces incorrect and made-up information, while the latter also produces errors.
- When tested by Motherboard, Grok produced misinformation about current events such as the Israel-Palestine conflict and the Pizzagate conspiracy theory. It also gave inaccurate timelines for news events.
- Grok's information is drawn from X posts, which could explain the inconsistencies and inaccuracies in its responses. Elon Musk has stated that users should expect rapid improvement in the chatbot.