Google announces its new big language model for AI

Google’s online office suite, Workspace, will soon allow users to generate entire documents and fill out spreadsheets with AI.

Google’s “Help Me Write” feature that can write essays and sales pitches will be included in Google Docs for everyone, the company said Wednesday. Another example the company gave was automatically writing job descriptions for online ads requesting a “senior job description for a textile designer.”

Google also demoed the software, based on PaLM 2, filling out an entire spreadsheet with three made-up dogs and notes for a fictional dog-walking company. Google says that you can take the structure and put in your own data and details.

Google Slides can generate AI images within the app, Google said. The example on stage created images of a slice of pizza dipped in fondue.

The new features will be available later this year to subscribers to a service called Duet AI, the company said. — How Leswing

Protesters wave a banner above Shoreline Amphitheater that reads “Protect Abortion Privacy,” according to CNBC reporters at the scene.

“Google has promised to remove location data about healthcare visits, including abortion care, but they STILL hold this sensitive data,” the protesters said. say on your website. — How Leswing

Google said it has built a new general-purpose large language model called PaLM 2 that can generate a wide variety of human-like text in response to human input. CNBC reported on Monday that Google planned to announce PaLM 2 during this week’s I/O conference.

The model is proficient in mathematics, software development, language translation reasoning, and natural language generation. “It’s better than our previous next-generation language models,” Zoubin Ghahramani, vice president of Google DeepMind, told reporters during a briefing earlier this week. Google announced the LaMDA model in 2021 and the most powerful Palm model in 2022.

PaLM 2 uses Google’s custom artificial intelligence chips and is more efficient to operate than previous models, Ghahramani said. The new model can work with more than 20 programming languages, including the old Fortran, and more than 100 spoken word languages, Ghahramani said.

About 20 Google products now use PaLM 2, he said, adding that a lightweight version can run on a mobile device.

Ghahramani declined to specify the exact sources he used to train PaLM or the number of parameters the model uses. OpenAI has taken a similar approach to disclosure with its GPT-4 model. —jordan novet

Many of Google’s top leaders are at I/O, including the company’s heads of human resources and search, CNBC reporter Jennifer Elias noted.

The new AI-powered functionality will obtain information about air quality, weather, and display a chosen route for Google Maps users. CEO Sundar Pichai demonstrated a Manhattan bike route that visualized a bike ride along the West Side Freeway, with AI-supported renderings of traffic, weather, and the route itself.

The demo suggested flyby functionality and that the feature would provide useful travel information overlaid on the user interface. Pichai said the functionality would roll out to more than a dozen major cities in the coming months.

—Rohan Gosvami

Google’s first big feature announcement could make it a lot easier to write emails, or it could fill your inbox with low-quality computer-generated messages.

Google will integrate generative AI into Gmail to help expand emails quickly, called “Help Me Type.” Google’s example was completing a letter to get a full refund from an airline. Google has settings to make the generated text longer or shorter, and uses the context of previous messages to help write the copy.

“Just type the notice of what you want, hit create, and a full draft will appear,” Pichai said. — How Leswing

The CEO of Google has taken the stage and is indicating that they will be talking about artificial intelligence all day.

“As you may have heard, AI has had a very busy year,” Pichai said.

“With generative AI, we are taking the next step,” he added.

Google is also likely to discuss updates to Android, which is the most widely used operating system in the world with 3 billion active devices. Android 14 is expected to launch sometime this summer and the company could spend some time talking about some of the new consumer features that will be available.

Google is also expected to introduce new hardware under its Pixel brand, such as the Pixel Fold, Pixel 7a, and a Pixel tablet. Google’s Pixel phones only have about 2% of the US market, according to Counterpoint Research. — As Leswing

Prior to the start of Google I/O, attendees will receive a free concert by independent electronic artist Dan Deacon.

Deacon’s song, “When I’m Done Dying,” was made partly with Google’s generative artificial intelligence tools, he said. The background was created by an image generator that reacted to the lyrics of the song.

“I’m going to play a few songs and a lot of the content will be created with Phenaki, Bard and MusicLM,” Deacon said. — How Leswing

This year, many of Google’s ads are expected to focus on artificial intelligence. AI has become the hottest part of the tech industry, partly due to the attention from OpenAI and its rival. Microsoftwhich has integrated so-called “big language models” that can generate flowing text and image generators that can generate original images into its Bing search engine.

But Google was heavily involved in creating these technologies, particularly the transformer architecture used by ChatGPT and other rivals. It has its own chatbot, Bard, which is currently limited to only a select number of beta testers. It also has one of the strongest AI departments, which was recently reorganized to become Google DeepMind. — As Leswing

We’re here at Google I/O 2023 sitting down and ready for the event to kick off.

The keynote should last a couple of hours and the announcements are likely to run the gamut from software news to hardware news. But, as we already said, artificial intelligence is likely to be the key focus of Google I/O 2023.

The company plans to announce a series of generative AI updates, including the release of a general-purpose large language model (LLM), CNBC informed on monday. –Jennifer Elias