Grok 3 Launches: Exciting Features Now Available to Select Users!

The highly anticipated Grok 3 AI model from Elon Musk’s AI company, xAI, is reportedly on the verge of release. Recently, users on the social media platform X have shared their experiences with this new model, which promises to enhance capabilities significantly compared to its predecessor, Grok 2.

What Users Are Saying About Grok 3

Over the weekend, several users, including notable reverse engineer Alexey Shabanov, reported gaining access to Grok 3 through the Grok chatbot app on X. Prior to losing access, users were able to test the model’s responses to various queries, showcasing its potential for logical reasoning and coding tasks.

Key Features of Grok 3

• Successfully answered complex riddles.
• Generated HTML and JavaScript code for a roulette wheel casino.
• Demonstrated advanced problem-solving skills.

However, it’s important to note that Grok 3 is still a work in progress, as it did encounter some issues, including minor programming errors in the roulette wheel code.

Unique Insights from Grok 3

Shabanov also revealed that Grok 3’s system prompt contains interesting directives, including the assertion that Donald Trump is the 47th president of the U.S. This could be seen as a measure to mitigate previous political biases in the AI’s responses.

Development Timeline and Expectations

Grok 3 has been under development for several months and was initially expected to launch last year. However, Musk recently indicated that the model has completed its critical pre-training phase, with a projected release date set for January or early February.

Training and Capabilities

xAI has utilized its powerful data center in Memphis, equipped with approximately 100,000 GPUs, to train Grok 3. Musk has claimed that this new model has been trained with ten times more computational power than Grok 2, making it a strong competitor against established models like OpenAI’s GPT-4 and Google’s Gemini.

Grok is designed not only to analyze images and respond to inquiries but also to potentially incorporate a “voice mode” for reading out responses, enhancing user interaction.

Controversial Aspects of Grok

When Musk first introduced Grok, he described it as an edgy and unfiltered alternative to existing AI systems, emphasizing its willingness to tackle controversial topics. While Grok does offer some unfiltered responses, it has shown a tendency to hedge on political issues, leaning towards the political left on subjects such as transgender rights and diversity.

Musk has attributed these tendencies to the training data sourced from public web pages and has expressed intentions to move Grok towards a more politically neutral stance. An “Unhinged Mode” is also in the works, promising responses that are intentionally provocative and controversial.

Future Updates and Enhancements

Additionally, Musk has announced that Grok 3’s training data will include legal documents, enhancing its understanding of legal topics. This could significantly broaden Grok 3’s applicability in various fields.

For ongoing updates about AI developments, consider subscribing to TechCrunch’s AI-focused newsletter, delivered every Wednesday.