The company says these improvements have improved Bard’s accuracy to computation-based word and math problems by 30%.

According to Google, the company is introducing a new technique called “implicit code execution” to help Bard detect computational prompts and run code in the background. The intended result is a more accurate response to mathematical tasks, coding questions and string manipulation prompts. These improvements also come with a new features that allows users to export a table to Google Sheets.

In a blog, Google leaders overseeing Bard say the improvements will make the generative AI chatbot better at answering questions such as:

• What are the prime factors of 15683615?
• Calculate the growth rate of my savings
• Reverse the word “Lollipop” for me

In the blog, Google says large language models (LLMs) are like prediction engines. Essentially, LLMs generate a response to prompts by predicting what words are likely to come next.

“As a result, they’ve been extremely capable on language and creative tasks, but weaker in areas like reasoning and math,” write Google Bard leaders. “In order to help solve more complex problems with advanced reasoning and logic capabilities, relying solely on LLM output isn’t enough.”

This new method, however, allows Bard to generate and execute code to boost its reasoning and math abilities.

According to Google, this approach is inspired from “a well-studied dichotomy in human intelligence, notably covered in Daniel Kahneman’s book “Thinking, Fast and Slow” — the separation of “System 1” and “System 2” thinking.

“System 1 thinking is fast, intuitive and effortless,” the Bard experts write. “When a jazz musician improvises on the spot or a touch-typer thinks about a word and watches it appear on the screen, they’re using System 1 thinking. System 2 thinking, by contrast, is slow, deliberate and effortful. When you’re carrying out long division or learning how to play an instrument, you’re using System 2.”

LLMs have been essentially operating under System 1, producing responses quickly but without deep thought, leading to some issues like trying to solve complex math problems.

Meanwhile, traditional computation more closely aligns with System 2 thinking as it is formulaic and flexible, but can produce impressive results with the “right sequence of steps,” Google says.

With the latest update, Google is combining the capabilities of both LLMs and traditional code – which it compared to combining System 1 and System 2 thinking.

“Through implicit code execution, Bard identifies prompts that might benefit from logical code, writes it “under the hood,” executes it and uses the result to generate a more accurate response,” Google says. “So far, we’ve seen this method improve the accuracy of Bard’s responses to computation-based word and math problems in our internal challenge datasets by approximately 30%.”

The post Google: Bard Now 30% Better at Computation-Based Problems appeared first on My TechDecisions.

The announcement came during the company’s annual Google I/O developer conference at which Google also announced new features for its generative AI chatbot Bard. Duet AI appears to be Google’s version of Microsoft’s generative AI assistant in Microsoft 365 which it calls Copilot.

Duet AI in Gmail, Slides, Sheets

According to Google, Duet AI already works behind the scenes in Workplace to help users refine emails and documents, and the company is now bringing the experience to Gmail on mobile to help users draft responses on the go.

The company’s initial launch of mobile will be fast-followed by contextual assistance that allows users to create professional replies that automatically fill in names and other relevant information.

This is in addion to Duet AI, Google says a new “Help me write” feature will be rolling out that allows users to leverage generative AI to help them write emails by typing a prompt of what they want to create a full draft of an email.

In Slides, Duet AI will help users generate images with a few words and generate an original visual to help support the text on the slides.

In Google Sheets, Duet AI will help users analyze and act on data with automated data classification and the creation of custom plans, the company says.

Google says classification tools understand the context of data in a cell and can assign a label to it to help eliminate manual data entry.

Also included in Sheets is a new “help me organize” capability that automatically creates custom plans for tasks, projects or any activity users want to track or manage. All they have to do is describe what they want to accomplish and Sheets will generate a plan for them, Google says.

Duet AI in Google Docs

In Docs, Duet AI will help users write content and include smart chips for information like location and status, as well as variables for details that users want to customize such. This adds to the “@ mention” feature in Docs that allows users to stay focused and collaborate in the document.

In addition, Google says it is adding new languages and bringing more capabilities to help with proofreading, tone and style in a new proofread suggestion pane that offers suggestions.

Duet AI in Google Meet

Like how Microsoft plans to bring Copilot into Microsoft Teams, Google wants to weave Duet AI into Google Meet, starting with the ability to generate unique backgrounds for video calls.

Test these features in Google Labs

To help users and organizations test out these new tools, Google is launching the waitlist for Google Labs and offering a limited number of spots to try out new Ai features in Search, Workspace, Project Tailwind and MusicLM.

Learn more about Google Labs and those projects in this short blog post.

The post Duet AI is Google’s Answer to Microsoft 365 Copilot appeared first on My TechDecisions.

According to Google, the Mountain View, Calif.-based tech giant’s updates include expanding access to Bard, bringing new visual capabilities to its AI-powered chatbot Bard, improving the generative AI’s coding capabilities and adding new productivity integrations.

Google Expanding Bard access

Purportedly in response to ChatGPT and competitor Microsoft’s massive investment in OpenAI, Google announced its large language model-powered chatbot Bard in February and launched it in March in a preview. Now, the company is removing the waitlist and opening up Bard to over 180 countries and territories, with more coming soon, Google says.

New Google Bard visual tools

In addition to new languages, Google is soon adding new visual interactions to Bard that will allow users to ask questions and get images in response to help explain the text. Also, users can use images in their own prompts via a Google Lens integration that will allow Bard to analyze the photo and include analysis in its response.

Google Bard Developer tools

Based on feedback from developers, Google is also introducing new features designed to make Bard a better coding assistant. With coding one of the main use cases of these early days of generative AI, Google hopes to strengthen Bard’s capabilities with more precise source citations, a dark theme and “export” button.

New Google Bard integrations

According to Google, export actions allow users to take Bard’s email and document drafts and move them directly into Gmail and Docs, respectively.

Source citations are rolling out next week, and the “export” button is listed as “coming soon.” Meanwhile, dark theme is now available.

The company also says it is working on integrating popular Google tools such as Docs, Drive, Gmail, Maps and more directly into Bard, and is also adding integrations with third-party software from Adobe, Kayak, OpenTable, ZipRecruiter, Instacart and more.

The post Google Bard Updates: New Coding Features, Integrations appeared first on My TechDecisions.

In a blog post, the company says it is starting to open access to Bard, a chatbot powered by a research large language model (LLM) and an optimized version of LaMDA, the company’s conversational AI technology. However, Google says it will update Bard with newer, more capable models over time.

According to Google, Bard is similar to other chatbots in that it is not always entirely accurate and has some limitations, but notes its “incredible benefits” such as assisting and facilitating productivity, creativity and curiosity.

When using Bard, users will often get multiple different drafts of responses from which to start. Users can continue to collaborate with Bard from there, including asking follow-up questions. Users can also simply ask Bard again for an alternative answer.

Similar to the new Microsoft Bing and its chat feature, Google calls bard a “direct interface to an LLM” and as a “complementary experience to Google Search.” Bard is designed so users can easily visit Search to check its responses or explore sources from across the web.

There will also be a “Google it” button to see suggestions for queries, and Search will open in a new tab so users can find relevant results and dig deeper, the company says.

“We’ll also be thoughtfully integrating LLMs into Search in a deeper way — more to come,” Google executives write in the blog, hinting at a potential chat component to Google Search, similar to how Microsoft integrated OpenAI’s newer ChatGPT model GPT-4 into Bing.

Google says it will continue to improve Bard and add new capabilities, including coding, more languages and multimodal experiences.

The launching of Bard comes more than a month after the company first announced Bard and when it fist began making Bard available to “trusted testers.”

In a blog post announcing Bard last month, Google CEO Sundar Pichai said external feedback will be combined with the company’s own internal testing to improve Bard and ensure quality and safety in real-world information.

The company’s chief executive detailed the company’s history of using AI to improve Google Search, including BERT, one of the company’s first Transformer models, and then MUM, which is 1,000 times more powerful than BERT.

“Now, our newest AI technologies — like LaMDA, PaLM, Imagen and MusicLM — are building on this, creating entirely new ways to engage with information, from language and images to video and audio,” Pichai says.

Google will now begin bringing these AI advancements to the users of its products, beginning with Search, Pichai writes.

Google’s announcement also comes the same day as Microsoft announced Image Creator in Bing, a new image generator running on an advanced version of OpenAI’s DALL-E.

The post Google Begins Making Its AI Chatbot Bard Available appeared first on My TechDecisions.

In a blog post, Google and parent company Alphabet CEO Sundar Pichai says Bard is powered by Google’s Language Model for Dialogue Applications, or LaMDA for short. Now, the company is making it more widely available, first as a lightweight model version of LaMDA that requires “significantly less computing power” and enables Google to scale to more users and solicit more feedback.

The company’s announcement is underpinned by several mentions of safety and responsibility, with Pichai touting Google’s set of AI Principles published in 2018.

What is Bard and when will it be available?

According to Pichai, Bard combines “the breadth of the world’s knowledge with the power, intelligence and creativity of our large language models,” drawing on information from the web to provide responses.

“Bard can be an outlet for creativity, and a launchpad for curiosity, helping you to explain new discoveries from NASA’s James Webb Space Telescope to a 9-year-old, or learn more about the best strikers in football right now, and then get drills to build your skills,” Pichai says writes in the blog.

Google is making Bard available to “trusted testers,” first and then they will make the AI more widely available to the public in the coming weeks.

External feedback will be combined with Google’s own internal testing to improve Bard and ensure quality and safety in real-world information, the Google CEO says.

“We’re excited for this phase of testing to help us continue to learn and improve Bard’s quality and speed,” Pichai writes.

The company’s chief executive details the company’s history of using AI to improve Google Search, including BERT, one of the company’s first Transformer models, and then MUM, which is 1,000 times more powerful than BERT.

“Now, our newest AI technologies — like LaMDA, PaLM, Imagen and MusicLM — are building on this, creating entirely new ways to engage with information, from language and images to video and audio,” Pichai says.

Google will now begin bringing these AI advancements to the users of its products, beginning with Search, Pichai writes.

What is coming to Search?

According to Pichai, Google Search users are increasingly searching for deeper insights and understanding rather than simple questions, but comprehensive learning about a complicated topic takes a lot of effort. AI can help synthesize and distill a breadth of information into easily digestible formats to help users grasp concepts more easily.

Soon, AI-powered features will be coming to Search to do just that, and help provide deeper understandings and additional perspectives when searching for information.

For developers

Starting next month, Google will start onboarding individual developers, creators and enterprises to try out the company’s Generative Language API, initially powered by LaMDA, with a range of models to follow.

The company plans to create a suite of tools and APIs designed to make it easy for others to build more innovative applications with AI. Google will also leverage its Google Cloud partnerships with Cohere, C3.ai and Anthropic to help scale these efforts and ensure enough compute power to build AI systems.

Why this matters

Several news reports on the internal discussions at Google since ChatGPT burst onto the scene–and since Microsoft’s reported $10 billion investment in parent company OpenAI–have suggested that Google is now intensely focused on building and touting its own AI models.

The company is planning to enhance Search with new AI features, which appears to be a direct response to rumors that Microsoft is planning to bring OpenAI’s technology to its own search engine, Bing.

Both companies have said they want to bring this advanced conversational AI to their products, and Microsoft is already starting with integrations into Azure and Teams.

The post Google Introduces Bard, Its Answer To ChatGPT appeared first on My TechDecisions.