Are we tokenmaxxing our way to nowhere?
Anthropic demos a secret, powerful model to the Fed Chair while OpenAI gobbles up companies.
The latest episode of TechCrunch's Equity podcast highlights a growing chasm between AI industry insiders and the broader public, marked by massive spending, strategic secrecy, and niche terminology like 'tokenmaxxing.' Hosts Kirsten Korosec, Anthony Ha, and Sean O'Kane point to OpenAI's aggressive acquisition spree—snapping up finance apps and media properties—and Anthropic's development of a model deemed too powerful for public release, yet demonstrated privately to Federal Reserve Chair Jerome Powell. This insider access underscores the concentrated power and opaque decision-making shaping the AI landscape.
Further illustrating the trend, the hosts discuss a notable shoe company's sudden pivot to rebranding itself as an AI infrastructure play, signaling the sector's overheated allure. The episode digs into the core battle for enterprise dominance between OpenAI and Anthropic, questioning what tangible infrastructure is being constructed beneath the hype. The analysis suggests that while capital and compute are being deployed at a staggering scale, the real-world applications and winners are still being determined, with access and influence increasingly concentrated among a select few.
- Anthropic has built an AI model considered too powerful for public release but demoed it to Fed Chair Jerome Powell.
- OpenAI is on an acquisition spree, buying companies ranging from finance apps to talk shows.
- A major shoe company has rebranded itself as an AI infrastructure company, highlighting the sector's hype.
Why It Matters
Concentrated AI development and insider access could dictate the technology's future, impacting markets and policy.