'Nvidia Way' Author Expects OpenAI To Counter Google, Anthropic With Much More Powerful Model In 'Coming Months'

Benzinga · 1d ago

Despite reports of an internal “Code Red” at OpenAI, the ‘Nvidia Way’ author Tae Kim predicts the ChatGPT maker is poised to retake the industry lead with a significantly more powerful model in the “coming months.”

The NVL72 Advantage For OpenAI

Speaking on CNBC International, Kim argued that the scaling laws driving artificial intelligence (AI) remain fully intact.

He forecasts that OpenAI will leverage Nvidia Corp.'s (NASDAQ:NVDA) newest hardware to counter recent surges by Alphabet Inc.’s (NASDAQ:GOOG) (NASDAQ:GOOGL) Google Gemini 3 and Anthropic's Claude, effectively answering the competitive crisis that has reportedly forced CEO Sam Altman to pause monetization efforts.

According to Kim, the recent performance leaps by competitors are actually bullish signals for the broader industry, proving that AI progress is accelerating rather than hitting a plateau.

He argues that OpenAI is currently positioned to benefit from a hardware advantage that its rivals did not have during their last training runs: Nvidia’s NVL72 server clusters.

“OpenAI is going to train on those new, more powerful clusters,” Kim said, noting that Microsoft Corp.‘s (NASDAQ:MSFT) data centers are now filling up with the new architecture.

“I fully expect in the coming months, OpenAI will come out with a much better performing model, just following the scaling laws that Anthropic and Google have done.”

A “Code Red” Environment

Kim's prediction comes at a pivotal moment for OpenAI. Internal leaks surfacing this week suggest the company has entered a “Code Red” operational state, freezing “side projects” like search ads and shopping agents to focus entirely on model quality.

With Google's Gemini 3 reportedly eroding ChatGPT's market share and user retention, the pressure is on OpenAI to prove that its next iteration can bridge the “reasoning deficit” currently plaguing its offerings.

Debunking the Bubble Narrative

Beyond the immediate model wars, Kim dismissed fears that the AI sector is approaching a dot-com style crash.

He noted that, unlike the year 2000, when market leaders traded at 100 times forward earnings, today’s major AI players are trading at a more grounded 25 to 30 times earnings.

“We are just at the beginning of a multi-year, multi-quarter acceleration of AI demand,” Kim stated, citing “overwhelming demand” that has forced hyperscalers like Microsoft and Amazon to double their data center capacity targets for the coming year.

Read Next:

Chamath Palihapitiya Sees Current Tech Giants Having An Upper Hand In AI Wars: ‘Google Has A Huge Runway’

Disclaimer: This content was partially produced with the help of AI tools and was reviewed and published by Benzinga editors.

Image via Shutterstock