Tuesday felt less like a tech update and more like a declaration of war. Or at least a desperate grab for the throne. Sundar Pichai took the stage at I/O 2026 and laid out a vision so aggressive it left the competition blinking. He called it the “agentic” era. Big words. He meant Google’s new AI won’t just answer questions anymore. It’ll do things for you. Autonomously.
Ten years in on their “AI-first” pivot, Pichai claims the mission remains the same: improve lives at scale. Sure. But the pressure is real. OpenAI, Microsoft, Anthropic—everyone is circling. Google had to show it still owns the space.
The Search Engine Gets a Brain Transplant
They tweaked the box you type into every day. The biggest change to Google Search in a quarter-century. That’s bold for a company that barely changes its homepage layout.
No more keywords. Not strictly. The new “Intelligent Search” lets you ramble. Type out your whole messy thought process.
“Context stays with you as you explore deeply.”
A demo showed a guy wanting pottery classes on Tuesdays. Just natural speech. No Boolean logic. No semicolons. Then, for heavier lifts—planning weddings, moving houses—Google introduces coding agents. It builds mini-apps right in the browser. Dashboards. Things you come back to.
Then there are the info agents. Set them and forget them. Ask Google to track your favorite athlete’s shoe collab, and it nags you when news hits. It crawls blogs, social feeds, everywhere.
It’s a response to the panic around “AI Overviews” killing web traffic. Now Google wants you inside their ecosystem doing work, not just clicking away to some indie blog. Smart play? Maybe. Annoying? Also yes.
Two New Gems
Or gemstones, depending on your material science degree. Google dropped Gemini Omni and Gemini 3.5.
Omni is the flashy one. Text, image, video, audio. It eats them all. You can edit a video just by talking to the AI. Tell it to ripple the mirror. Turn an arm to metal. It understands context now—history, science, culture mixed into one reasoning engine.
One demo featured a hand touching a mirror that liquefied. Cool trick. But it shows the model understands what “next” looks like in a physical scene.
The other side? Gemini 3.5 Flash. Cheap. Fast. Built for devs and auditors who hate burning cash. Google says it’s half the price of other frontier models and does the job in a fraction of the time. Then they teased 3.5 Pro for later this year. Because of course they did.
Apps That Finally Talk Back
The bleed-over into productivity tools is complete now.
Gmail lets you search by voice. Docs Live creates documents from your muttered notes. YouTube added “Ask.” You can finally query a video without scrubbing through twenty minutes of intro music to find the one fact you need.
The video jumps straight to the relevant part.
For coders, Android Studio can generate apps from natural language. Tell it what you want in English. Get Java code back.
And the hardware? Android XR smart glasses returned to the fold. Not headsets. Actual glasses. Audio versions give you spoken help—directions, translations, photo editing. You say “Hey Google” or tap the frame.
Display glasses? Live info overlaid on the world.
Google partnered with Warby Parker and Gentle Monster. Fashion meets tech, hopefully without the headache. The audio glasses drop this fall.
Will anyone actually wear them? Time will tell. The ecosystem is tight, the features are flashy, but the barrier to entry for smart eyewear has always been social stigma. Pichai thinks AI agents can break through that. He might be right. He might be chasing ghosts.
