Elon Musk’s pioneering artificial intelligence startup xAI has unveiled its latest Grok-1.5 chatbot. The enhanced version, which will soon be rolling out to users on the X platform (formerly Twitter), boasts improved reasoning capabilities.
In an announcement on his social media platform X, Musk said the new chatbot will be available to early testers and existing Grok users on X next week. “Grok 2 should exceed current AI on all metrics. In training now,” Musk added.
Should be available on 𝕏 next week.
Grok 2 should exceed current AI on all metrics. In training now. https://t.co/s7rVxvK8mZ
— Elon Musk (@elonmusk) March 29, 2024
Grok-1.5 represents a substantial upgrade from its predecessor, boasting enhanced reasoning abilities and an impressive context length of 128,000 tokens, setting a new benchmark in large language models.
“Grok-1.5 comes with improved reasoning capabilities and a context length of 128,000 tokens. Available on soon,” xAI said in a blog post.
Powered by an open-source large language model, Grok-1.5 also showcases remarkable proficiency in coding and mathematics, achieving remarkable scores across various benchmarks. Impressively, it boasts stellar performances on MATH (50.6%), GSM8K (90%), and HumanEval (74.1%), positioning itself as a formidable contender in the AI arena.
“One of the most notable improvements in Grok-1.5 is its performance in coding and math-related tasks. In our tests, Grok-1.5 achieved a 50.6% score on the MATH benchmark and a 90% score on the GSM8K benchmark, two math benchmarks covering a wide range of grade school to high school competition problems. Additionally, it scored 74.1% on the HumanEval benchmark, which evaluates code generation and problem-solving abilities.”
Grok 1.5 compared to other LLMs
Speculation is rife that Grok-1.5 might introduce an innovative “Analysis” feature, capable of summarizing extensive discussions and threads within the X platform while streamlining user interactions and enhancing overall engagement.
While benchmark results hint at Grok-1.5’s parity with the esteemed GPT-4, its true prowess lies in its capacity to harness real-time data streams on X. Recent discussions between Grok and industry titan Midjourney hint at a potential alliance, signaling xAI’s ambitious foray into the burgeoning AI chatbot market.
Meanwhile, this announcement follows Musk’s bold decision to open-source Grok, a strategic move aimed at democratizing AI technologies. It comes hot on the heels of legal skirmishes with OpenAI, underscoring Musk’s unwavering commitment to fostering innovation in the face of industry competition.
The launch of Grok-1.5 marks a pivotal moment in the evolution of conversational AI. With xAI, Musk said he aims to provide an alternative to the status quo dominated by OpenAI and Google. Musk’s vision for AI safety, articulated at Britain’s AI Safety Summit last year, includes the establishment of a “third-party referee” to monitor AI development and raise red flags when necessary.
Launched last year, xAI promises to be a beacon of “maximum truth-seeking AI.” Grok’s rollout for Premium+ subscribers earlier signaled Musk’s intent to disrupt the AI landscape, a sentiment echoed in his podcast conversation with Lex Fridman, where he lamented OpenAI’s departure from its nonprofit, open-source roots.
Musk first announced the launch of Grok on November 6, 2023, adding that the AI chatbot will possess a “rebellious streak” and bring a sense of humor and sarcasm, which sets it apart from rival chatbots like ChatGPT and Google Bard.
