Today, that hesitation looks like the moment Google let someone else seize its future. OpenAI did exactly what Google chose not to: it put a chatbot front and centre. When ChatGPT launched in late 2022, it didn’t just captivate the world — it exposed how slowly Google had moved to commercialise its own breakthroughs, including the transformer architecture OpenAI freely borrowed from Google’s 2017 research.
But as ChatGPT marks its third anniversary, the pendulum is swinging back.
According to BloombergAlphabet’s resurgence is so strong that its stock is now within striking distance of Nvidia’s $4.2 trillion valuation, propelled by rave reviews for its Gemini 3 model and growing demand for its homegrown tensor processing units (TPUs).
Google has spent the past three years building out a full-stack AI ecosystem largely behind the curtain — from chips and cloud infrastructure to frontier models and consumer apps. The payoff is now visible everywhere: a revitalised cloud business, surging demand for its homegrown AI chips, and a new generation of Gemini models that are winning industry respect and rattling competitors from OpenAI to Nvidia.
Also Read: ChatGPT effect: In three years AI chatbot has changed way people look things up
The Gemini 3 moment
Google’s turning point came with the release of Gemini 3. The model earned immediate praise for its reasoning, coding abilities and performance on niche tasks that typically stump chatbots. It quickly climbed AI leaderboards, prompting Andrej Karpathy — one of OpenAI’s founding members — to call it “clearly a tier 1 (Large Language Model) LLM.”For Google, it was vindication. A company once accused of moving too slowly was suddenly matching OpenAI’s pace.
The chip play no one saw coming
What makes Google’s ascent more striking is how deeply vertical its AI bets have become.
Google is one of the only companies on the planet that produces the entire AI stack: the apps (like the Nano Banana image generator), the foundational models, the cloud infrastructure and the chips underneath. Demand for its TPUs — once used almost exclusively by Google itself — is now rising.
AI startup Anthropic struck a deal to use up to 1 million Google TPUs, a contract worth tens of billions. Meta is reportedly exploring Google’s chips for its data centres from 2027, a development that sent Alphabet’s valuation soaring and triggered a selloff in SoftBank (which backs OpenAI) and a dip in Nvidia shares.
As Ben Barringer of Quilter Cheviot told Bloomberg“Many others have failed in their quest to build custom chips, but Google can clearly add another string to its bow here.”
Wall Street has moved on from the “Google is behind” narrative
Alphabet has added nearly $1 trillion in market cap since mid-October, buoyed by Warren Buffett’s $4.9 billion stake and growing conviction that Google won’t cede the AI future to OpenAI, Microsoft or anyone else.
Neil Shah of Counterpoint Research summed it up best, saying that Google has always been “a sleeping giant” and it’s suddenly awake.
A company built for the long game
Google’s confidence has long stemmed from its economics. It has:
- A massive, constantly refreshed data corpus
- A profitable core business that funds moonshot R&D
- Large-scale computing infrastructure
Also Read: Is Google-Alphabet becoming the new leader of AI stocks as Gemini 3 gains power?
A unified, full-stack approach to AI
CEO Sundar Pichai put it plainly on an earnings call: “We’ve taken a full, deep, full-stack approach to AI. And that really plays out.”
That approach accelerated after a 2023 reorganisation that placed DeepMind’s Demis Hassabis at the helm of Google’s combined AI division. Hassabis’s mandate: focus almost entirely on frontier models and pull ahead. He has reportedly kept key talent from defecting, despite multimillion-dollar offers from rivals.
Google’s meteoric market rally and the warnings
Alphabet’s rally has pushed its valuation beyond historic norms. Its forward earnings multiple has risen to 27x — above its 10-year average of 20x, Bloomberg noted. Its 14-day Relative Strength Index (RSI) has touched 75, indicating the stock may be overbought.
Matthew Maley of Miller Tabak told Bloomberg that Google “could take the leadership role for the AI industry going forward,” but cautioned that the stock “could be due for a pullback.”
The competition isn’t standing still
OpenAI still leads in consumer mindshare, though. ChatGPT recently hit 800 million weekly users, while Google says 650 million use its Gemini app. And OpenAI and Microsoft have deeper enterprise traction than Google Cloud, which remains third behind AWS and Azure.
However, Bloomberg notes that Google’s AI momentum is undeniable, even if its enterprise adoption lags. As Thomas Husson of Forrester told the publication: “It’s definitely fair to say that Google is back in the game with Gemini 3.”
To paraphrase Mark Twain, he added, reports of Google’s death “have been widely exaggerated.”
Also Read: Google’s AI infra spend triples to over $90 billion as Pichai plays down bubble fears
Three years in, OpenAI is no longer alone
As ChatGPT completes three years, the AI landscape looks very different from the one OpenAI dominated in 2022. Back then, Google was scrambling to respond to a product inspired by its own research. Today, Google has quietly turned its research depth, infrastructure and chip strategy into a formidable challenge to OpenAI’s first-mover advantage.
And as much as OpenAI pushed the industry into its generative AI era, it’s Google that spent the past three years assembling the machine that could define what comes next.