Google I/O 2023: Bard is now available for chat, amidst intensive AI focus

It is a good thing you play a game that involved gulping down an alcoholic shot every time AI was muttered at Google I/O 2023 keynote. It might have all gone pear shaped, because the tech giant’s focus on AI is so exhaustive, it spans everything from the Bard chatbot, a new search tool to identify AI generated images, a coding bot for Android developers, an AI tool for app developers to help them build Play Store listings and Android’s generative AI dreams for text messages and wallpapers. You get the idea.

Google Bard Chatbot: Globally available, in limited languages

There is no doubt that Google has kicked off this summer’s developer conference chapter, setting quite a benchmark for the likes of Microsoft, Meta, Amazon and indeed Apple to try and match in the coming weeks. The biggest announcement is the availability of the Bard chatbot, albeit with limited languages, for everyone who is willing to try it. This means you finally complete the troika of chatbots in your life, with Bard joining OpenAI’s ChatGPT and Microsoft’s Bing chatbot.

But it is clear Google is not willing to take any risks with user safety or the chatbot misbehaving (misinformation, bias etc.) when interacting with a user. Till now, Bard was accessible to a very small set of users, in the UK and US. The rest, had to join a waitlist. This is the first major expansion of Bard AI’s availability, on a global scale.

“As we continue to make additional improvements and introduce new features, we want to get Bard into more people’s hands so they can try it out and share their feedback with us,” says Sissie Hsiao, Vice President and General Manager for Google Assistant and Bard. This means Bard is now available in 180 countries, with support for English, Korean and Japanese.

Hsiao confirms they are on track to add support for 40 more languages, in the coming weeks. “As we further expand, we’ll continue to maintain our high standards for quality and local nuances while also ensuring we adhere to our AI Principles,” she adds.

Languages aren’t the only upcoming additions to look for. Bard will soon be able to have a conversation based on an image you may have added to the chat and will also generate search results with more visuals.

Google is looking to plug in Bard into its own suite of apps, including Docs, Drive, Gmail and Maps. There will be third-party integrations as well, including with Adobe Firefly to help users generate high quality AI images.

Generative AI: The future of Android?

Uncharacteristically, Android was not the focus for Google at this year’s iteration of the annual developer conference (it usually is, and quite elaborate too) – though there was talk about Android on foldables and tablets, as well as Android TV and smartwatches. But we did get a glimpse of the AI infusion expected with Android 14, that’s likely to roll out later this summer.

Android’s Messages app will get something called Magic Compose. This will generate reply suggestions “based on the context of your messages” with attempts made to make them sound as you may usually do in a text conversation. The beta version rolls out sometime in the next few weeks.

“Make your messages more concise, professional or even written like Shakespeare himself! Experiment for yourself and try Magic Compose when it rolls out this summer in beta,” says Dave Burke, Vice President of Engineering at Google.

There will also be two visual features, called Cinematic Wallpaper and Generative AI Wallpaper. The former will use on-device machine learning to add a 3D dimension to your favorite gallery photos, which can then be used as the wallpaper on the phone. The latter will use Google’s text-to-image diffusion model to generate an image, which again you will be able to use as the phone’s wallpaper.

Think of this as Midjourney or Stable Diffusion, but for a very specific functionality on your Android phone. It is not clear if these images, generated by either functionality, will be shareable with other users or devices.

AI for Search: Distinguish between AI images and real photos

In the coming weeks, Google will add a new information layer in search results, which will provide users with more context about the image – when that image was first indexed on Google Search, similar images and their indexing timelines as well as the websites they have appeared on.

This comes at a time when AI generated images, which are looking increasingly real and accurate, have caught many off guard. In the last few months, there have been instances of fake images doing rounds on social media, including one seemingly capturing the moment former US President Donald Trump is arrested. AI tools such as Stable Diffusion and Midjourney are becoming increasingly accurate with detailing, including body features and skin tones.

“With this background information on an image, you can get a better understanding of whether an image is reliable — or if you need to take a second look. For example, with About this image, you’d be able to see that news articles pointed out that this image depicting a staged moon landing was AI-generated,” says Cory Dunton, Product Manager for Search, referring to an AI generated image of a moon landing.

AI for developers: coding and listing help

There will be a new AI bot that’ll help app developers with coding. It’ll be called Studio Bot and is expected to help developers build apps by generating code and fixing errors. Google says the bot is built on Codey, a new foundational coding model, an evolution of the updated PaLM 2 large language model (LLM). The Studio Bot supports Kotlin and Java programming languages.

App developers who may often face the conundrum of getting the perfect app description ready for the Play Store listing page, will soon be able to take help from AI. Google has built a bot that’ll help do the work that otherwise would have required a human to write a perfect copy. This tool is now available for developers, albeit only in the English language for now.