GoogleS Gemini 3 Leapfrogs OpenAI, Igniting New AI Pressure Cooker
OpenAI, the frontrunner in the artificial intelligence revolution, faces its most significant challenge since the launch of ChatGPT, as Google and Anthropic rapidly close the gap in cutting-edge AI technology. Three years after captivating the world with its popular chatbot, the $500 billion start-up is grappling with escalating data center costs, the complexities of maintaining its technological edge, and the ongoing struggle to retain top talent.
The pressure intensified last week with the release of Google’s Gemini 3, a large language model widely considered to have surpassed OpenAI’s GPT-5 and achieved advancements in model training that have eluded its competitor in recent months. “Its quite a strong difference with the world we had two years ago where openai was leading ahead of everyone else,” said Thomas Wolf, co-founder and chief science officer of open-source start-up Hugging Face. “It’s a new world.”
Even before Gemini 3’s debut, OpenAI’s chief Sam Altman warned employees in a memo last month to prepare for “short-term competitive pressure” and “rough vibes,” as reported by The Facts. This internal acknowledgment underscores the shifting dynamics in the AI landscape.
A year ago, many observers had dismissed Google’s attempts to catch up to OpenAI.Concerns that ChatGPT and other AI-powered search applications like Perplexity would cannibalize Google’s core search business led to a significant lag in Alphabet’s stock price compared to other Big Tech companies during the AI-driven market rally of 2023 and 2024. However, Google’s fortunes began to turn earlier this year, fueled by a series of confident updates at its I/O developer conference in May and the viral success of its Nano Banana AI photo-editing tool this summer. These developments boosted monthly users of the Gemini mobile app to 650 million, a substantial increase from approximately 400 million in May.
Alphabet’s shares have surged in recent months, pushing its market capitalization toward $4 trillion for the first time, reflecting Wall Street’s growing confidence in Google’s ability to leverage its dominance in search, cloud infrastructure, and smartphones to deliver new AI capabilities to billions of users. Koray Kavukcuoglu, Google’s AI architect and DeepMind’s chief technology officer, told the Financial Times that the company had “pushed our performance quite significantly” by training its AI models using Google’s own custom chips. “Being able to connect with consumers, customers, companies, at that scale is really something that we can do as of that full stack integrated approach that we have,” he added.
That “full stack” approach includes Google’s custom tensor processing unit (TPU) chips, which allow the company to train Gemini 3 without relying on the expensive Nvidia chips used by most of the AI industry. “I think we have a unique approach there,” said
,founded in 2021 by former OpenAI staffers,is currently raising a new funding round expected to value the company at over $300 billion. While Anthropic’s Claude chatbot has been overshadowed by ChatGPT’s consumer success, its focus on AI safety has resulted in a more reliable tool for corporate clients, according to its supporters, and its coding tools are widely regarded as best-in-class.
Despite OpenAI’s continued dominance with over 800 million weekly chatbot users, data from web analytics company Similarweb indicates that people are now spending more time interacting with Gemini than ChatGPT.The launch of Gemini 3 also propelled Google’s AI app higher in the US and UK iPhone app store rankings, although ChatGPT remains the top AI app according to Sensor Tower.
Erik Brynjolfsson, author and professor at the Stanford institute for Human-Centered AI, cautioned against prematurely counting OpenAI out, citing its diverse range of new applications as a potential source of revenue to fund its core research. “All these companies have a surplus of very profitable opportunities all around them,” he said. “There’s room for multiple companies to do extremely well because the possibility is so large.”
