Google Adds AI To Its AI Using AI

In partnership with

Nearly a decade ago, Sundar Pichai announced to shareholders that Google was shifting from being a mobile-first company to being an AI-first company.

The announcement felt premature.

The vernacular de jour in 2016 was “machine learning”.

Sure, AI-adjacent things were happening.

The company released TensorFlow that year, a free, open-source software library for machine learning and artificial intelligence projects.

DeepMind, acquired by Google and left initially to its own devices, built a machine that could beat some of the best players in the world at a game of Go that spring.

Impressive achievements for sure, but also nerdy and mundane.

Where was the AI?

And who cared?

Then, two Googlers, Ashish Vaswani and Jakob Uszkoreit, had a hallway conversation about how they could improve natural language translation.

They collaborated with six other Google researchers, some of whom worked in Geoff Hinton’s Google Brain lab in Toronto, on a paper called “Attention Is All You Need.”

These fine folk gave us the Transformer as well as a series of other methods that are the cornerstone of what we now refer to as AI.

One of the paper’s authors explained it like this, “It’s a general method that captures interactions between pieces in a sentence, or the notes in music, or pixels in an image, or parts of a protein. It can be purposed for any task.”

Around Christmas that year, a Canadian Googler who was working on expanding translation in Mountain View came to visit the Toronto office.

We had visitor desks in our pod. I guess our Real Estate and Workplace Services team thought that the automotive ad sales folks were outgoing enough to chat with anyone.

They were right.

We befriended this engineer during his trip to visit his family for the holidays.

He explained how he was using AI to accelerate Google’s translation efforts. They now only need partial human translation of a language to extract and extrapolate the rest by using AI.

It was a little mind-blowing.

It was also the first time I had heard of an application of AI or machine learning technology that wasn’t just about improving how we served ads or how we found cost savings by learning to cool data centers differently.

Although Google Research got a rebrand in 2018 to “Google AI”, the rest of the company wouldn’t adjust its machine learning communication strategy until after it got poked in the eye by Sam Altman and OpenAI’s ChatGPT in 2022.

The memo came internally in 2023. We no longer discuss machine learning; it’s now called “Google AI.”

Today, Google is drunk on AI.

At the 2025 Google IO event, they announced how they were jamming AI into all of their products. Additionally, “AI” was mentioned 92 times; it was practically its own AI drinking game.

My Pixel 10 Pro XL is littered with AI features (most of which I haven’t even begun to explore, yet). As WIRED reported, this includes “real-time translated phone calls in your own AI-generated voice, an AI camera coach that makes you a better Instagram husband, and a voicemail replacement.”

Google Search’s English version of AI Mode is now available in over 180 new countries and territories. Their AI translation efforts will likely enable that service in additional languages soon enough.

Google has made its most powerful AI video creator, Veo 3, available for everyone to use on its Vertex AI platform.

Google’s Gemini has popped in ranking in the App Store thanks to a new AI image model, Nano Banana.

This week, Google Chrome, having survived the US Department of Justice’s chopping block, announced it is jumping with both feet into the 2025 browser wars, getting its own Gemini AI, “agentic” cursor-controlling tools and likely a whole host of other AI treatments in the future.

Also this week, Google is making it possible to share your own Gemini Gems, custom AI assistants and experts designed for specific tasks.

Google continues to eat its own dog food. Sundar proclaimed last year that 25% of Google’s newest code is being written by (or with) AI. That might have been an embellishment.

But it’s clear that Google is adding AI to its AI using AI.

Even in Google’s ad ecosystem, if you want to run your creative in Google’s AI Mode or AI Overviews, they have already suggested that you’ll only be able to do that if you’re buying AI-based ad formats like Performance Max, AI Max or letting loose the dogs of search with Broad Match.

At a time when Tim Cook and Sam Altman are travelling the UK to have dinner with The Donald.

And Mark Zuckerberg is spending a lot of money to supercharge his own AI team while his AI-powered Ray-Bans (codenamed Hypernova) fail him during a demo.

Sundar and Google march forward, late to start but not at all last in the AI race.

Turn AI Into Your Income Stream

The AI economy is booming, and smart entrepreneurs are already profiting. Subscribe to Mindstream and get instant access to 200+ proven strategies to monetize AI tools like ChatGPT, Midjourney, and more. From content creation to automation services, discover actionable ways to build your AI-powered income. No coding required, just practical strategies that work.

Subscribe to Get Your Free Guide