OpenAI Leads Superintelligence Race, Anthropic Accelerates with Claude 4, Meta Poaches Top AI Talent

AI analyst Dylan Patel of SemiAnalysis asserts that OpenAI is positioned to be the first to achieve superintelligence, citing their consistent leadership in foundational AI breakthroughs, particularly in reasoning capabilities. This assessment places OpenAI at the forefront of the highly competitive race to develop advanced artificial intelligence.

Patel ranks Anthropic as a strong second, noting a recent acceleration in their development efforts. Previously known for a cautious approach focused on AI safety, Anthropic has demonstrated increased momentum with significant product launches, including the recent release of Claude 4. This new model family, comprising Claude Opus 4 and Claude Sonnet 4, introduces hybrid reasoning modes, extended thinking capabilities, and enhanced coding performance, with Opus 4 achieving a 72.5% success rate on the SWE-bench coding benchmark.

The competition for third place is described as a tight contest among Google, xAI, and Meta. Meta, in particular, has intensified its pursuit of superintelligence through aggressive talent acquisition. The company reportedly made offers exceeding $100 million to top AI researchers and acquired Scale AI primarily to secure its leadership, including CEO Alexander Wang, to head Meta's new Super Intelligence Labs (MSL).

Mark Zuckerberg, Meta's CEO, has publicly shifted his focus, now emphasizing the pursuit of superintelligence as a primary goal for the company. This strategic pivot has led to the formation of MSL, which has successfully recruited key researchers from rival firms like OpenAI and Google DeepMind. OpenAI Chief Scientist Mark Chen acknowledged a "brain drain" to Meta, prompting OpenAI to recalibrate compensation to retain its talent.

Despite Meta's aggressive recruitment and Anthropic's recent advancements, Patel maintains that OpenAI's track record of innovation, including pioneering reasoning models, gives them a distinct advantage. The race for superintelligence continues to be a high-stakes endeavor, with leading AI firms investing heavily in talent and computational resources to gain a decisive edge.