Google I/O, Google’s annual tech conference/showcase was held yesterday at the Shoreline Amphitheater in Mountain View, California. 

Fresh off of Alphabet’s (NASDAQ: GOOGL) Q1 earnings call, where the company reported revenues exceeding expectations, many people tuned into Google I/O to see how the tech giant planned to keep its momentum.

Here are the highlights…

New and Improved Pixels

Google made their mobile device line the key focus of their conference, announcing three additions to the Pixel product line.

Most notably, the Pixel Fold, which will be Google’s first foldable mobile device, which will mirror the $1800 price tag of the Samsung Galaxy Z Fold4, the most popular phone of its kind.

Also like the Fold4, Google’s new offering will feature a 7.6-inch OLED screen.

The Pixel Fold is also notable in that it will be the thinnest mobile device ever offered by Google.

When it’s not folded in half, of course.

Also included in the new Pixel lineup are the more affordably priced Pixel 7A and 11-inch Pixel Tablet.

Improvement Through AI

While Google may have tried to make their conference’s main focus their mobile line, it’s what they presented in the world of artificial intelligence that has everyone’s attention and has Alphabet’s stock up over 5% in the last 24 hours.

Headlining their AI announcements is that Google Bard is now widely available, no longer requiring users to sign up for a waitlist.

While only available in English, Google plans to offer the AI personal assistant in 40 more languages in the near future.

Bard’s first trial run was a rocky one, answering user questions with responses that were easily proven to be incorrect.

It seems, in offering Bard to the general public, that Google would have to be confident in their improvements.

However they do include a humble disclaimer in Bard’s greeting to new users…

While it’s commonly thought that Bard is a ChatGPT clone, and an inferior one at that, Google claims that the newly upgraded chatbot is especially good at tackling coding queries. Making it a powerful tool for debugging and explaining existing code.

Google clarifies that Bard is not a replacement for Google Search.

Search, however, is getting a similar boost thanks to AI.

Soon, if users choose to opt into the “Search Generative Experience (SGE)”, they will start seeing AI Snapshots.

Snapshots are intended to augment Google’s trademark “10 Blue Links” result format by including a naturally worded and presented answer generated by their artificial intelligence algorithm.

Searches can be refined by answering AI-generated followup questions.

These improvements to both Bard and Search are driven by Google’s newest Large Language Model (LLM) dubbed PaLM 2, which was also unveiled at Google I/O.

CEO Sundar Pichai said of PaLM 2, which is already being used to power all of Google’s AI services…

“PaLM 2 models are stronger in logic and reasoning, thanks to broad training in logic and reasoning. It’s also trained on multilingual text spanning over 100 languages.”

Senior research director Slav Petrov showed off PaLM 2 by showing it clear a hurdle that has tripped up most other AI chatbots…

Understanding idioms.

Here you can see a translation from German to English that avoids a possible misunderstanding caused by being overly literal:

PaLM 2-powered AI enhancements are expected to be used to improve most of Google’s products in the coming years.

Filling Out The Lineup

AI certainly stole the show in Mountain View, also powering new photo editing features, an improved Workspace, and Android customization.

However, there were also other notable announcements.

Google Home, the smart home control center app, is now widely available to the public.

Perspectives will be a search feature that focuses on human-generated answers and discussions sourced from Reddit, Stack Overflow, and blogs.

Google also announced that the Wear OS 4 will be available for smartwatches later this year.

With that, we want to hear your reactions. How do you feel about Google’s mobile lineup? Do you think their AI improvements are more than just playing catch-up with OpenAI and Microsoft?