All the AI announcements from Google I/O 2023

Google had a lot to announce at this year’s Google I/O event. And most of it was in some way AI-related.

The sudden explosion of artificial intelligence products on the scene led by OpenAI’s popular ChatGPT has clearly shaped what Google’s been working on. Don’t take Mashable’s word for it. Google CEO Sundar Pichai said so himself. Pichai opened up the event by sharing that AI has had a very big year and said Google has now positioned itself as an “AI-first” company. In doing so, Google has “reimagined all of its core products.”

Gmail will write emails for you

Google I/O 2023 kicked off with a focus on its most popular non-search related products: Gmail, Google Maps, and Google Photos.

First up, Gmail is expanding on its Smart Reply and Smart Compose features, which help users auto-complete emails by recommending replies and sentences. The new AI-powered Help Me Write feature will allow users to auto generate entire emails. Users can even decide to expand on short AI-created content in order to include more details. The feature will even pull relevant details from previous related email correspondences with the recipient.

AI-powered, ‘immersive’ Google Maps

Google Maps’ upcoming AI rollout includes a feature called Immersive View for Routes. The feature transports users into a digitally created model of what their exact route will look like. Users can even travel along the route in order to spot landmarks and stops along the way. This feature is reportedly coming to 15 cities by the end of the year.

Google I/O: Photos
Credit: Mashable Screenshot

All sorts of AI wizardry in Google Photos

Google Photos is adding to its AI-powered features such as Find photos and Magic Eraser with Magic Editor. The standout moments in the demo for Magic Editor include the ability to move individual elements in a photo and an impressive feature that auto-generates elements that are cut-off by the photo in order to make them whole if a user wants to center them in a photo. 

Google searches will soon produce AI answers

Google’s core product also has a slew of AI features on the way.

Google I/O: Search
Credit: Mashable Screenshot

Google Search will soon respond to users’ search queries in a conversational way and, perhaps most impressively, find web results for multifaceted queries instead of having to split each into an individual search.

Additional follow-up search query questions will also be suggested by Google Search once a user shows interest in a particular link or product.

Google I/O: Search
Credit: Mashable Screenshot

The user interface for AI-powered Google Search is pretty clunky looking though, although Google made clear that it’s an evolving work in progress.

All of these Google Search AI features will still return search results powered by Google’s popular search ranking system which has made the company’s search engine the undisputed leader over the past few decades.

Google I/O: Misinformation
Credit: Mashable Screenshot

In a short, interesting segment, the company showcased some of its developments it’s working on to tackle misinformation being spread via AI. One example showed how Google Images and Google Lens will be able to track the origin of AI-generated images.

New Bard features, powered by a new language model

Perhaps the most anticipated news from the Google I/O event is in regards to a brand new Google product: it’s AI chatbot Bard, which is now powered by Google’s newest large language model: PaLM 2(opens in a new tab). PaLM 2 is an upgrade to the language model Google released in April of 2022(opens in a new tab), PaLM. The new model is reportedly fluent in over 100 languages including multiple types of code, and has received math, logic, and reasoning upgrades as well.

At the event, Google announced it was dropping the waitlist for Bard, which will open the floodgates and allow anyone to access the AI chatbot. Also, Bard is on track to supporting 40 new languages as Google adds Japanese and Korean to the English-based product.

Bard has receive kind of a bad wrap in recent months, with many Ai aficionados feeling like Google’s chatbot falls short of the more popular ChatGPT. It’s hard to tell from a live product demo but Google showed off Bard writing code, creating photo captions, and even creating AI-generated images with Adobe Firefly via Bard’s upcoming extensions feature which will allow the chatbot to integrate with third-party apps and platforms. And it all worked just fine during the presentation.

AI sprinkled into all the products in Google Workspace

Generative AI features will soon be coming to Google Workspace products. While Google didn’t go through them all, we did get some interesting demos showcasing a few examples.

Google I/O: Workspace
Credit: Mashable Screenshot

Users will be able to utilize AI to create full-on spreadsheet templates in Google Sheets and even insert data into the corresponding columns automatically. In Google Slides, users can spice up their slides with auto-generated AI art based on the content of the slide, without any additional prompts. And a “contextual collaboration” feature, called Duet AI for Workspace, is on the way, and will allow users to collaborate with others in Documents, Sheets, and more with an assist from AI.

Google did have a big new non-AI product to announce, the Pixel Fold — a foldable Google Pixel phone. However, at a starting price of $1,799, the Pixel Fold is probably out of reach for most consumers shopping for a smartphone.

Google I/O: Pixel Fold
Credit: Mashable Screenshot

All in all, Google’s Android product line felt like an afterthought compared to the company’s new focus on being an “AI-first” company. And in stark contrast to the vertigo-inducing price point of the Pixel Fold, it appears that all of those new AI features are free to use, at least for now.