The Most Important AI Developments from Google I/O


Google just made some huge AI announcements at Google I/O—some of which could have big implications for marketers and business leaders.

They include:

  • Gemini 1.5 Flash, a new model that Google says is just as powerful as 1.5 Pro, but much faster and more efficient.
  • Gemini 1.5 Pro’s context window has doubled to 2 million tokens. Gemini 1.5 Pro with 1 million tokens (the current context window) will be available to Gemini Advanced users.
  • Related to context windows, CEO Sundar Pichai said that Google is “on a journey towards the ultimate goal of infinite context.”
  • Gemini 1.5 Pro will be infused into all Google Workspace apps sometime in the (hopefully) near future, which means it will be able to find and use files across Drive, summarize and search emails, and help you be more productive across Docs, Sheets, Slides, and Gmail.
  • Google announced Project Astra, AI that can see and understand the world around you using your phone’s camera. The company’s demos of Astra show it answering questions about what it sees and remembering object locations.
  • Google Search is getting a controversial overhaul as AI Overviews (formerly Search Generative Experience) are rolling out to all users in the US for select searches. These are AI-generated summaries of certain types of search results that appear before the traditional list of blue links in SERPs.
  • Google is releasing Gems, a way to create custom chatbots in Gemini (just like OpenAI’s GPTs). 
  • Veo is the company’s answer to OpenAI’s Sora video generator. Veo generates 1080p video from text, image, and video prompts.
  • The company announced its newest TPU (tensor processing unit) chip called Trillium, which provides a 4.7X improvement in compute over Google’s previous version of chips.

It’s important to note that most of the features are announcements, not releases.

While AI Overviews have started to roll out, most of the other developments didn’t come with firm timelines for when we all get access to them.

That doesn’t make them any less exciting. It just means you might have to wait for them.

So, which announcements matter most?

I got the answer from Marketing AI Institute founder and CEO Paul Roetzer on Episode 98 of The Artificial Intelligence Show.

Gemini + Workspace = Winning

Roetzer is “extremely excited” about Gemini’s increasing infusion into Google Workspace. 

Imagine having an AI that can recall almost anything across your documents, emails, and internal company knowledge to help you complete complex tasks.

“In the next 12 to 18 months, it’s going to start to become very commonplace where people are going to just have Google with Gemini connected to their knowledge bases,” predicts Roetzer. 

As an added bonus, Gemini infused into Workspace also means Gmail and Drive search might no longer, well, suck.

“This has always been one of the more perplexing things in my professional life—how useless Gmail search and Google Drive search are,” says Roetzer. “They’re just painfully bad search functions. And yet this is the greatest search company on Earth.”

Gemini could change all of that.

Context is king

Alongside the announcement of Gemini Pro’s 2 million token context window, CEO Sundar Pichai said that Google is “on a journey towards the ultimate goal of infinite context.”

This matters, says Roetzer.

“That was the first time I’ve heard them talk about infinite context, which basically means memory, the ability of these things to not forget and to very accurately find things within [the context you give them].”

That unlocks a whole range of capabilities these models haven’t previously had. While context windows have been expanding, the models themselves have not historically excelled at recalling and retrieving all of the information they’ve been asked to remember with a high degree of accuracy.

That looks like it’s starting to change.

“As this context window keeps getting bigger, it’s going to enable this instant recall of anything that’s ever happened,” says Roetzer.

AI in the physical world

“The big potential one is certainly Project Astra,” says Roetzer. This is AI that can understand everything it sees through your camera.

The Astra demos showed off a capability called “Live,” which lets users have real-time voice chats with Astra about their environment.

All of this points to a future powered by AI assistants that can see, understand and interact with the physical world to help you.

It could also signal a future where Astra gets baked into wearable hardware—something Google has dabbled with in the past.

“I wouldn’t be surprised at all if we don’t hear something related to Google Glasses being resurrected here at some point,” says Roetzer.

We need to talk about Search

A big immediate question that came out of Google I/O was:

With the rollout of AI Overviews, is this the end of Google Search as we know it?

“I think people are rightfully concerned about this,” says Roetzer. “There’s going to be an impact to organic search.”

How big an impact remains to be seen.

Google hasn’t shared much data on engagement with AI Overviews yet. And they don’t appear for all types of search results.

But Roetzer advises preparing for a traffic drop, even if it doesn’t happen immediately. His advice? 

Diversify your content distribution. Publish content across social media, email newsletters, and other channels. That way, you’re not dependent on search traffic alone.

“Plan for the fact that maybe your search traffic drops in a significant way. And don’t wait for that to happen before you start diversifying your strategy,” he says.

The future will be here sooner than you think

Google’s announcements paint a picture of an AI-powered future that is coming soon to a screen near you. While the updates aren’t all available yet, the writing is on the wall.

Gemini’s integration with Workspace brings the power of huge language models to common productivity tools. Project Astra and smart glasses prototypes point to a world filled with AI agents helping you navigate daily life.

Even Google’s under-the-radar technical improvements in chips, architectures and context windows foretell big leaps in multimodal AI for business use cases.

Search remains a big question mark. Traffic drops are a real possibility for brands. Forward-thinking companies will diversify their content strategies to be prepared.

Says Roetzer:

“Again, not all things we’re going to see today are going to affect our jobs tomorrow. But it definitely hints at where it’s going.”



Leave a Reply

Your email address will not be published. Required fields are marked *