In partnership with
This newsletter is a script edit of our latest YouTube video, which you can watch now.
There’s an awful lot of noise out there. If you’ve been opening this newsletter in the last few months, you’ll have noticed that we don’t really get the whole ‘GEO’ thing in SEO (Generative Engine Optimization). We particularly don’t like silly graphs showing ChatGPT traffic overtaking Google soon. Indeed, that’s because looking at the real data, it’s clear Google has an almighty lead.
Code red
But there’s a back story. In December 2022, Google declared a “Code Red” - an internal alarm reserved for existential threats. ChatGPT had just launched, racing to 100 million users faster than any tech product in history (at the time).
ChatGPT’s launch led to alarms sounding at the search giant.
Fast forward three years, and the roles appear to have reversed. OpenAI is scrambling to respond to Google’s rapid progress with Gemini 3.0, pushing out GPT-5.2 at breakneck speed. Google, once seen as slow and cautious, suddenly looks like it has momentum. The question now being asked seriously is not whether OpenAI disrupted Google - but whether Google is now overtaking OpenAI.
Public perception still lags reality. OpenAI releases are greeted with breathless hype, while major Google upgrades often land with a thud. Part of this is narrative. Google is the incumbent, the verb, the monopoly people love to root against. OpenAI, led by the charismatic Sam Altman, plays the role of the insurgent hero. Everyone loves an underdog story, after all.
AI assistants certainly feel like the future of information retrieval. Traditional search - clicking through ad-stuffed pages to find an answer - already feels dated. Chatbots provide faster, cleaner responses, and hallucinations have become less frequent. Entire industries, particularly SEO, are in a panic, with regularly suggestions that ChatGPT will soon replace Google as the primary gateway to information.
Shoppers are adding to cart for the holidays
Over the next year, Roku predicts that 100% of the streaming audience will see ads. For growth marketers in 2026, CTV will remain an important “safe space” as AI creates widespread disruption in the search and social channels. Plus, easier access to self-serve CTV ad buying tools and targeting options will lead to a surge in locally-targeted streaming campaigns.
Read our guide to find out why growth marketers should make sure CTV is part of their 2026 media mix.
Google’s scale advantage
But this suggestion flies in the face of Google’s HUGE scale advantage:
Google processes roughly 13 billion searches per day and is by far the most visited domain on the Internet - with 130 billion sessions a month.
YouTube, also owned by Google, serves 70 billion sessions a month and hosts over 20 billion videos, an unmatched source of multimodal training data.
Google Chrome controls about 65% of the global browser market.
Android runs on more than 70% of the world’s smartphones.
Google Workspace has 3 billion monthly users.
ChatGPT, impressive as it is, operates at a completely different order of magnitude. It has about 800m weekly users and 5 billion sessions a month.
This matters because the AI race is not really about chatbots - it’s about who reaches artificial general intelligence (AGI) first. And AGI is fundamentally a game of compute, data, distribution, and capital.
Compute and benchmarks
On compute, Google has a quiet but enormous advantage. For over a decade it has been building its own AI chips, Tensor Processing Units (TPUs), rather than relying on Nvidia GPUs. These chips are cheaper to run at scale and allow Google to train larger models, run more experiments, and iterate faster. The fact that companies like Meta and Anthropic are now using Google’s TPU infrastructure tells you how competitive it has become.
Google’s TPUs give it a distinct compute advantage.
Benchmarks, while imperfect, also suggest momentum. Gemini 3 has topped major leaderboards and outperformed GPT-5 on difficult reasoning tests. Scale AI founder Alexandr Wang recently described Gemini as “without a doubt the leading model.”
Leadership and money
More strikingly, Geoffrey Hinton - the “Godfather of AI” and former Google researcher - has said he believes Google will win the AI race precisely because of its compute advantage.
Then there’s money. Alphabet generates over $100 billion in quarterly revenue and tens of billions in profit. OpenAI, by contrast, is reportedly burning around $8 billion per year, with cumulative losses projected well into the next decade. Google can afford to play a long, grinding game. OpenAI must continually justify itself to investors.
Leadership styles also matter. OpenAI’s story is inseparable from Sam Altman, whose ability to build alliances and momentum has carried the company through repeated crises - including the extraordinary 2023 boardroom coup that briefly saw him fired and reinstated within days. Google, by contrast, has no single visionary figurehead driving AI. It is a machine: slower, less glamorous, but more resilient.
A guaranteed victory?
But none of this guarantees a Google victory. Benchmarks can be gamed. This isn’t a two-player race - Anthropic, Meta, xAI, Microsoft, and China all matter.
ChatGPT still dominates consumer mindshare and daily usage, which counts for more than raw model quality. And history shows that incumbents can still stumble.
But the parallels are hard to ignore. Google once entered a crowded search market late and crushed faster-moving pioneers like AltaVista and Ask Jeeves through superior technology, distribution, and business fundamentals. If AI becomes as central as search, Google is structurally positioned to do the same again.
Google invented the transformer, watched others commercialise it first, panicked - and now appears to be calmly clawing its way back. When Geoffrey Hinton says Google is “beginning to overtake” OpenAI, it’s worth paying attention. This race is far from over, but the era of OpenAI’s uncontested lead may already be ending.
