xAI’s Grok 3 Declared ‘World’s Smartest AI’ Amid Betting Market Support, Experts Offer Mixed Reactions

xAI’s Grok 3 Declared ‘World’s Smartest AI’ Amid Betting Market Support, Experts Offer Mixed Reactions

This is not investment advice. The author holds no positions in the stocks mentioned herein.

xAI Unveils Grok 3 LLM: A Game Changer or Overhyped?

In an exciting reveal, xAI launched its Grok 3 language model (LLM) during a live stream on Monday, hosted by none other than Elon Musk. The AI company has branded Grok 3 as an unparalleled advancement in artificial intelligence; however, several industry experts are casting doubt on its advertised benchmarks, citing notable shortcomings.

According to a post from xAI, the Grok 3 model is being touted as the “world’s smartest AI, ”sparking intrigue across various sectors.

A complete video of the demonstration can be found in the post linked above. Additionally, following what has been dubbed the “DeepSeek effect, ”Musk announced that the earlier version, Grok 2, will soon be open-sourced, offering further insights into the technology’s development.

xAI has been proactive in asserting that Grok 3 surpasses all other publicly available foundational models, such as DeepSeek-V3 and GPT-4o, particularly in areas like mathematics, science, and programming. The LLM even achieved an impressive score of 1, 402 on the Arena benchmark.

Meanwhile, in the world of speculative investing, a betting contract on Manifold Markets regarding Grok 3 being crowned as the most powerful AI is leaning toward a “yes”conclusion. However, we observe a notable drop in the probability from 91% late Monday night to just 78% currently.

It appears that the emerging critical reviews of Grok 3, though limited, may be influencing these declining probabilities.

For instance, Zihan Wang, a former DeepSeek employee, posed a physics question to Grok 3 by presenting an image of two iron balls of different sizes suspended at different heights from the Leaning Tower of Pisa, asking which would hit the ground first. The expected logical answer would be the heavier ball, yet Grok 3 incorrectly stated that both would land simultaneously.

Moreover, there are growing questions surrounding xAI’s decision not to release Grok 3’s performance metrics on established benchmarks like FrontierMath, Arc-AGI, or HLE.

It’s important to note that these critiques are not intended to diminish Grok 3’s potential, which is undoubtedly a formidable AI model. Rather, they raise important questions about the authenticity of xAI’s claims regarding its superiority.

Financial Developments and Future Prospects

In a separate yet equally important development, Bloomberg recently reported that xAI is seeking up to $10 billion in new funding, potentially catapulting its valuation to $75 billion. Previously, the startup secured $6 billion during a funding round that valued it at $40 billion.

It’s worth noting that Guodong Zhang from xAI announced that Grok 3 was trained using an impressive 100, 000 GPUs, indicating a significant leap in resources and capabilities. This development comes amidst predictions that the revenue from AI chip sales could soar to $227 billion by 2032.

For more details and insights, you can check the full article here.

Leave a Reply

Your email address will not be published. Required fields are marked *