Grok-3 Launch Sparks AI Competition as OpenAI Prepares GPT-4.5

Is it fear, uncertainty, and doubt, or just another twist in the escalating AI Race? Elon Musk is set to reveal Grok-3 tonight, testing the extensive computing power of xAI against major competitors like OpenAI and Anthropic.

Musk has called Grok-3 "scary smart," and his announcement on X has seemingly triggered a response from his adversary, OpenAI CEO Sam Altman. Altman stated that his upcoming model, GPT-4.5, is close to achieving artificial general intelligence (AGI). The real question is how soon this model will actually be released.

Recently, Altman challenged Musk to simply focus on creating a superior product. "Just compete by building a better product," he told Bloomberg. Musk seems ready to take up that challenge.

In his promotional build-up, Musk claims Grok-3 is "outperforming anything that has been released so far." The live demonstration is scheduled for Monday night at 8 PM PT.

"Smartest AI on Earth." — Elon Musk (@elonmusk) February 16, 2025

Analysts and enthusiasts in the AI community are contemplating whether Musk's claims hold any weight, especially considering that xAI's previous models have fallen short of expectations. If Grok-3 lives up to the hype, it could significantly shake up the AI landscape—unless, of course, OpenAI's GPT-4.5 debuts first.

Mark Teneholtz, head of AI at the forecasting platform Predelo, noted, "We’re about to see what a skilled team can achieve with a significantly larger training cluster than what was used for GPT-4." He was referring to the Colossus Supercluster, which consists of 100,000 Nvidia HGX H100 GPUs used to train xAI's latest models. Originally, Altman hinted that GPT-4.5 could arrive within weeks, but his recent comments imply there may be a quicker release in an attempt to outpace Musk.

With already heightened concerns about competitors like China's DeepSeek and significant financial stakes at play, OpenAI cannot afford to be idle. When a Twitter user suggested that Altman should launch a surprise attack on Grok-3, Altman quipped that "that wouldn't be very nice."

The model GPT-4.5, codenamed Orion, represents OpenAI's final "non-chain-of-thought" offering before the company consolidates its AI technology in GPT-5. This iteration has been trained using synthetic data from the o1 reasoning model, further contributing to the AGI-like capabilities that Altman has alluded to. Interestingly, Grok-3 is also being promoted as exceptionally skilled in reasoning.

However, skepticism remains regarding Musk's claims. Benjamin De Kraker, a former xAI employee, resigned after the company pressured him to retract a post where he ranked Grok-3 below ChatGPT-o1 and ChatGPT-o3 in coding capabilities. De Kraker clarified that his opinions led to a backlash from xAI leadership, despite Musk labeling the situation as "weird."

"I resigned from xAI tonight. It makes me very sad, but it was the right thing to do -- and here's why. xAI told me I either had to delete the post quoted below, or face being fired," he stated.

Charles Hoskinson, the creator of Cardano, quickly reached out to offer him a position.

A well-known leaker on X, known as "@Iruletheworld," claimed that Grok-3 would leave people in awe, asserting that it represents AGI. The leaker expressed, "It doesn’t guess. It doesn't fumble logic. It doesn't need training wheels. It thinks. It reasons. It plans. When you talk to it, you feel a weight of genuine intelligence staring back at you. This is what happens when you push scale to the limit and stop being cautious."

xAI employee Frederik Meringdal even suggested to the leaker that Grok-3 might be capable enough to prompt him to change his profile description, asserting that Grok-3 surpasses AGI and achieves Artificial Super Intelligence.

This rivalry between Musk and Altman is steeped in history. After co-founding OpenAI in 2015, Musk left in 2018 after his attempt to merge OpenAI with Tesla was rejected. Earlier this year, Musk's investor group tried unsuccessfully to acquire OpenAI for $97.4 billion, a bid that Altman dismissed. Musk has stated he would withdraw if OpenAI reverted to its original non-profit mission.

OpenAI reflects that Musk's departure was due to his failed attempt at leading as CEO. Their dispute has since escalated into legal action, with Musk suing OpenAI over its partnership with Microsoft, claiming it strays from the initial non-profit mission of the organization.

It seems that the future of AI could hinge on whether either of their companies achieves AGI or Super Intelligence, potentially settling the ongoing conflict between these tech titans.

AI, rivalry, Grok-3