Google’s AI is learning faster than your flatmate cramming for finals

GOOGLE

Google’s AI is learning faster than your flatmate cramming for finals

Google’s I/O 2025 was all about one thing: weaving AI into everything.

From Search and Gemini to creative tools and hardware, nearly every product is getting a smarter upgrade.

Search is now more than a search engine.

The new “AI Mode” lets you chat back and forth, use your camera to ask questions in real time, track prices on products, and even try on clothes virtually.

You’ll also get support for analysing data, booking appointments, and summarising sports or finance info, all inside Search.

Gemini, Google’s main AI model, has levelled up.

Gemini 2.5 and the faster Gemini Flash now offer better reasoning, more secure tool use, and smarter research features.

You can upload PDFs, connect your Gmail or Drive, and let Gemini build reports, diagrams, or summaries from both public and personal sources.

A new Agent Mode is being tested, too, where you simply describe what you need, and Gemini gets it done.

A rival for ChatGPT Image and Midjourney?

On the creative front, tools like Imagen 4 and Veo 3 are now live.

Imagen sharpens image quality and adds support for spelling and typography, while Veo can generate full videos with audio.

Google’s new Flow tool lets anyone create cinematic films with just a prompt, while Lyria RealTime gives you AI-powered music creation that’s interactive and expressive.

There are new ways to access it all.

Google AI Ultra, their top-tier subscription, includes 30TB of storage, YouTube Premium, and full access to Gemini Pro and Flash.

Students in the UK and other countries can get a free Gemini upgrade for the full academic year.

Meanwhile, Google AI Pro offers core features for £19.99/month.

Google also previewed their XR smart glasses, which are being tested with live translation, messaging, and turn-by-turn directions powered by Gemini.

The Beam platform (evolving from Project Starline) will bring more immersive video calls to offices this year.

Here’s what to keep in mind:

• Search and Gemini are now more visual, personalised, and capable of full tasks.

• Imagen, Veo, and Flow make it easier than ever to generate video, music and design content.

• Google is gradually turning AI into the core layer of every product, including future smart glasses.

Vids, vibes, and voiceovers

And if you’re a developer, there’s a lot to dig into: Jules helps manage codebases, Stitch builds UI designs from prompts, and Gemini now powers tools like Firebase, Android Studio, and the new open-source models MedGemma and SignGemma for medical and accessibility-focused apps.

Even Gmail is getting AI-smart replies based on your own tone and past messages. NotebookLM will soon support video summaries of long documents, and Sparkify, a new Labs experiment, turns questions into animated explainer videos.

Imagine telling your calculator from 2005 that Search can now see your dinner and book it a table.