Grok-3 is the latest advanced AI chatbot developed by xAI, Elon Musk's AI. Launched today, Grok-3 boasts over ten times the computational power of its predecessor, Grok-2, and introduces enhanced reasoning capabilities designed to tackle complex tasks by breaking them into smaller components and self-verifying solutions before responding.
In early testing, Grok-3 has demonstrated superior performance compared to models like OpenAI's GPT-4o, Google's Gemini, and DeepSeek's V3. It offers two distinct reasoning modes: "Think," which displays Grok's thought process during problem-solving, and "Big Brain," intended for more computationally intensive tasks. Additionally, xAI has introduced Deep Search, a next-generation AI search engine, similar to the deep search agents of Perplexity, Gemini, and ChatGPT. A synthesized voice feature for Grok is rumored to be coming in the near future.
Access to Grok-3's functionalities is available through the X Premium Plus subscription, which recently increased in price ($40 per month), with an option for an advanced SuperGrok plan. Despite aiming for maximized truth-seeking capabilities, previous versions faced criticism for misinformation and offensive outputs. xAI plans to open-source Grok-2 in the near future.
I asked Perplexity to help me come up with 5 prompts that would test Grok-3. One of the reasons I test chatbots is to determine how reliable they are, interestingly enough, after noticing Grok-3 did not always site sources, I had to tweak the prompts to ensure I would be able to do my own research to fact-check the chatbot.\
To read more, click here.