Elon Musk has recently launched a new product named Grok through his latest AI startup, xAI. This new chatbot is designed to compete directly with OpenAI’s ChatGPT and other generative AI chatbots.
Musk has touted Grok as being superior in certain aspects, claiming it to be “the best there is”. The chatbot has reportedly outperformed GPT 3.5 in various academic tests, including those related to mathematics and coding. Grok operates on a novel engine called Grok-1 and has access to real-time information from X.
In a series of evaluations using standard machine-learning benchmarks, Grok surpassed all other models in its compute class, including ChatGPT-3.5 and Inflection-1. These benchmarks were designed to measure math and reasoning abilities. However, Grok was only outdone by models that had been trained with significantly larger amounts of training data and computing resources, such as GPT-4.
xAI also evaluated Grok, Claude-2, and GPT-4 on the 2023 Hungarian national high school finals in mathematics. Grok and Claude-2 both achieved a C grade with scores of 59% and 55% respectively, while GPT-4 received a B grade with a score of 68%.
Musk has also revealed that Grok is designed to respond with a touch of sarcasm and humor, and will answer provocative questions that most other AI systems would reject.
Despite its impressive performance, Grok is still in its early beta stage and is improving rapidly every week. xAI is working on implementing more safety measures to prevent malicious use of Grok. Currently, Grok is only available to a limited number of users in the United States and is offered as part of the X Premium Plus subscription for $16 per month.