Google Used ChatGPT to Improve Bard, Scale AI Documents Reveal

“So we’re going to start off with saying when Bard interacts with Gmail, Drive and Docs, it’s only when a user has opted in to say it’s okay,” he says. Scale AI never mentioned Google by name in the documents, referring instead to its anonymous “client.” It references Bard over a dozen times in a private Google sheet titled “bard rewrite comparison with gpt4,” and a slide in one training document includes Google’s logo. Unlike Microsoft’s Bing chatbot, Bard doesn’t have footnotes with web sources.

Google’s answer to ChatGPT: Bard. Here’s what you need to know about its new AI chatbot.

There’s also a “Google it” button that will turn your prompt into a more search-engine-friendly query and direct it to Google Search. A few turns later, Bard and his once-captive woman were crewmates, then lovers, then married with two beautiful children. For the most part, it’s tough to get Bard to say something truly wild.

You can use the three-dot menu button on the bottom-right to copy the response to your clipboard, to paste elsewhere. And finally, you can modify your question with the edit button in the top-right. But if the AI is unsure, the sentence may be highlighted in orange to indicate that it knows this part of the answer might be wrong.

Google’s Bard chatbot can now tap into your Google apps, double-check answers and more

This should help users better understand when the AI is “hallucinating” — that is, when it provides a response based on false information. This is a problem with modern AI that may confidently generate output even when it doesn’t have the supporting data. Scale AI also used ChatGPT to improve Bard’s responses in specific domains, like engineering or physics. In an update from August 2023, Scale AI managers wrote that they would have staff “redo” Google’s AI answers for engineering-related questions “with GPT4 guidance.” Scale AI is a San Francisco startup that does crucial AI grunt work for Big Tech.

Google gives the example of submitting a photo of your dogs alongside the prompt “write a funny caption about these two.” Google Lens identifies the breeds of the dogs, and Bard then writes something relevant to their characteristics. It’s a bit of a tricky feature to unpack, but it could have a lot of creative potential — depending on how well the system is integrated. Google’s chatbot is supposed to be able to explain complex subjects such as outer space discoveries in terms simple enough for a child to understand.

OpenAI’s terms of service at the time prohibited others from using its output “to develop models that compete with OpenAI.” Scale AI and Google did not respond to a question about whether they got permission from OpenAI for these detailed comparisons and rewrites. Google is updating its Bard AI chatbot to help developers write and debug code. Rivals like ChatGPT and Bing AI have supported code generation, but Google says it has been “one of the top requests” it has received since opening up access to Bard last month. The long-term vision for all these products, once we all finish trying to make them say racist things and tell fart jokes, is to build a general AI that can help us manage all phases of our life. Right now, the best we can hope for is half-decent information retrieval and quasi-well-written blog posts, and Bing seems to do both better.

• Other than the Apprentice Bard, Google is also testing other AI-powered products, including a search page.
• The search giant said in a blog post that teens can use the tool to “find inspiration, discover new hobbies and solve everyday problems.” Teens can ask Bard about important topics, like which universities to apply to, or about more fun topics, like how to learn a new sport.
• However, like ChatGPT, Google’s AI technology isn’t fully there yet — responses may be inaccurate or even offensive, according to Google.
• Microsoft on Tuesday announced that it would bring Bing Image Creator – a tool that uses AI to turn text prompts into images –  to the new AI-powered Bing and Edge preview.
• When it launched earlier that year, Bard, which has since been renamed Gemini, was internally mocked as “rushed” and “botched.”

You might ask Bard to give you tips to reach your goal of reading more books this year, explain quantum physics in simple terms or spark your creativity by outlining a blog post,” Google VP of Product Sissie Hsiao and Google VP of Research Eli Collins wrote in a blog post. For more general uses, Google is making Bard more visual, with the ability to analyze images, offer images in query results, and generate visuals using AI (a feature that’s arriving “in the coming months” powered by Adobe’s Firefly software). Google launched an artificial intelligence chatbot called Bard on Tuesday, the company’s response to the AI frenzy dominating big tech companies after the launch of OpenAI’s ChatGPT—backed by Google’s rival Microsoft—caught the attention of millions last year. Today’s announcement comes a few weeks after Google opened up its generative AI search experience to teenagers. The AI-powered search experience, also known as SGE (Search Generative Experience), introduces a conversational mode to Google Search where you can ask Google questions about a topic in a conversational manner.

Google’s Bard AI chatbot can now help you code and create functions for Google Sheets

Scale AI told BI that the ChatGPT outputs weren’t used to train Google’s or any others’ models and were part of routine “evaluations,” which it said are industry standards. In other words, today’s limited release of Bard is the first step of a long process. It’s going to be interesting to see how people interact with the product when the product becomes more widely available, and how regulators and content creators feel about Google’s new product.