Google I/O is an annual event for developers organized by Google. During the conference, Google introduces the latest goods and services and gives a preview of what's coming. 2023 is a year to witness the leap of the AI industry, and Google I/O 2023, the company’s 15th annual developer conference, is also very AI-related.
This year’s Google I/O event is over. Let's look at what Google has announced:
1. Mobile Devices
Pixel Fold
To achieve success in the market for foldable smartphones, Google introduced the Pixel Fold. The 7.6" OLED display of the Pixel Fold is shown when it opens and folds like a book, just like Samsung's Galaxy Z Fold 4 (which debuted in August 2022). It is also thinner than the Galaxy Z Fold 4 and has a 5.8-inch external display. The Pixel Fold, on the other hand, is Google's entry into the foldable smartphone market and will face off against comparable goods from competitor firms like Samsung, Vivo, Oppo, and Xiaomi.
Pixel 7a
The Pixel 7A, a brand-new member of Google's midrange A-series family, has been unveiled as a more affordable option to the Pixel 7 smartphone line. The Pixel 7a can be a wise investment in the mid-range smartphone segment when compared to the $1,799 price tag of the Pixel Fold. Its 6.1-inch display is smaller than the Pixel 7's 6.4-inch one, making it similar to the Pixel 6a.
Pixel Tablet
The tech giant has also launched the Pixel Pad, a new tablet powered by the same Tensor G2 engine as the Pixel series, whose main feature includes the Charging Speaker Dock. The Pixel tablet supports the USI 2.0 stylus. The company's return to the tablet industry was also marked by the introduction of the Pixel Tablet.
The new Google Tensor G2 (5 nm) chipset powers the Pixel Fold, Pixel 7a, and Pixel Table.
2. Google's Advanced AI Features
PaLM 2
PaLM (Pathways Language Model), the large language model (LLM) that Google introduced last year, has undergone an update with PaLM 2. PaLM 2 has four models: Gecko, Otter, Bison, and Unicorn. These four models serve as the foundation for advanced AI features across 25 Google services and products, including Workspace (Gmail, Google Docs, etc.) and Bard.
According to Google, PaLM 2 will be able to solve math problems on its own. Additionally, PaLM is multilingual and can generate text in over 100 other languages, including essays, emails, idioms, poems, puzzles, and more. According to Google, PaLM 2 is competent in 20 of the most widely used programming languages. Python, JavaScript, Prolog, Fortran, Verilog, and other programming languages are all supported by PaLM 2.
According to Frederic, PaLM 2 will power Google's updated Bard chat tool and function as the foundation model for most of the new AI features. "This is viewed as an effort by Google to aim at its major rival, OpenAI's GPT-4”, he added.
Smarter Bard
Google demonstrated AI features in search (similar to a ChatGPT-integrated Bing) when it released Bard, the generative AI chatbot, in February 2023.
The PaLM 2 LLM now powers Bard itself. Bard used to have a waitlist that wasn't available everywhere, but today it is available in over 180 countries and will continue to expand. It implies that anyone can use the generative AI tool across 180 nations.
In the future, Google will make it possible to combine Bard with its services, such as Docs, Drive, Gmail, and Maps. To encourage innovation, Google, for instance, has linked Adobe Firefly and Bard.
The capability of Bard to add images to its comments is a further improvement. Aside from AI-generated photos, Bard is also capable of writing code, coming up with descriptions for pictures, and other similar content creation activities.
3. Google AI-enhanced products and services
Play Store
Developers can optimize Android apps, write app listings content effortlessly and summarize app reviews
Magic Editor and Magic Compose
Google is bringing MagicEditor to Photos to deliver an enhanced editing experience. MagicEditor will be available on a few Pixel phones. The Magic Editor can remove redundant elements (passersby, redundant objects, etc.) and move the subject of the photo to a better position with the background of the photo.
Another feature is Google Messages' Magic Compose, which allows AI to help users compose messages in a personal voice, which is effective for those who often have to text.
Live View and Immersive View
Google Maps could become more visual with Live View and Immersive View for routes. These features will aggregate information and simulate traffic, lanes, complex intersections, parking areas, etc. You can even see the weather in that area.
In addition, Google is also trying to create 3D activities that allow users to experience simulated tourist destinations, including local tourism brands, to enhance the travel experience. Google will be testing AI-upgraded Maps in 15 cities, including San Francisco, Los Angeles, New York, Las Vegas, and Miami.
Codey
There is also a new tool specifically for code completion support that will power PaLM 2.
Google Workspace
Google's workspace includes Gmail, Calendar, Drive, Docs, Meet,... Now all these tools are integrated with AI to make the job easier.
You can completely create a Google Sheets table with just a few descriptions and create images for visual slides with the drawing tool in Google Slides. In Google Meets, you can completely create your own personalized wallpaper.
MusicLM
Convert text to music. MusicLM is a deepfake platform. Note that the music generated by this tool should be used for personal use only. Because there were a lot of music companies suing these tracks, they won.
Google Search
2 new features for Google image search are expected to be added: telling searchers "where is an AI-generated image" and adding a function called "about the image" that provides information like the first export present, appearing on any website or any social media.
Sidekick
Sidekick is a new tool that was announced by Google and is integrated into Google Docs. The tool will show up as a hint. Google has tried writing a fairy tale on Docs, and Sidekick has shown quite a few suggestions for writing a sequel to the story.
Imagen
This is a text-to-image tool that was announced last year. The improvement is being able to edit photos and add captions.
Project Tailwind
Google introduced Project Tailwind, an AI note-taking utility that pulls data from users' drives, providing content summaries
The way it works is that you can simply select files from Google Drive, and it will effectively create a personalized and private AI model with expert knowledge of the information you give it.
Wear OS 4
Google has launched the next version of its smartwatch operating system, Wear OS 4. Battery life and also newly developed and improved accessibility features, such as text-to-speech.
Universal Translator
This is a new tool in the video. After translating into another language, it will use AI to change the speaker's lips to match the translated content.
Health Connect
Health Connect is Google's health data storage platform. Some of the cool features coming soon include new exercise routes and an easier way to record your menstrual cycle. Additionally, MyFitnessPal now has integration with Health Connect, allowing users with type 1 and type 2 diabetes to access glucose data in the MyFitnessPal app.
