Google is now testing Bard, its answer to OpenAI’s ChatGPT. We also have gotten our first look at AI-powered features in Google Search.
Google CEO Sundar Pichai today announced the soft launch of Bard to a set “trusted testers,” showcasing what this early experimental release looks like with some examples.
We knew this was coming because Pichai mentioned it on the latest Google earnings call.
What Google’s AI features will look like in Search
Let’s first start with what ChatGPT-style AI will look like in Google Search – because that’s what everyone has been waiting to see.
Here is a screenshot Google provided. It shows how Googe Search uses AI to respond to long queries with a long AI-created response:
This screenshot gives you an idea of how Google could soon integrate AI-generated answers directly into Google Search.
“Soon, you’ll see AI-powered features in Search that distill complex information and multiple perspectives into easy-to-digest formats, so you can quickly understand the big picture and learn more from the web: whether that’s seeking out additional perspectives, like blogs from people who play both piano and guitar, or going deeper on a related topic, like steps to get started as a beginner.”
And to be clear: Bard is separate from Google search. The above screenshot is highlighting the AI-powered features that will be rolling out on Google Search “soon.”
What Bard looks like
Here is another screenshot from Google showing how Bard, its standalone ChatGPT-style service, might respond to a question:
In this case, Bard is explaining new discoveries from NASA’s James Webb Space Telescope to a 9-year-old.
As you can see, there is no one right answer to this question.
What is Google Bard?
Google explains that Bard is an experimental conversational AI service, powered by LaMDA (Language Model for Dialogue Applications).
LaMDA was announced two years ago, at the 2021 Google I/O keynote, as Google’s next-generation language and conversation capabilities. It is now being used to create these answers for users.
Bard is powered by a lightweight model version of LaMDA, which requires “significantly less computing power,” Google said. Google said this will allow them to scale to more users while also allowing for more feedback.
Pichai also wrote:
“Bard seeks to combine the breadth of the world’s knowledge with the power, intelligence, and creativity of our large language models. It draws on information from the web to provide fresh, high-quality 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.”
Bard is experimental
Google emphasized that Bard is experimental and they are excited for this “phase of testing to help us continue to learn and improve Bard’s quality and speed.”
Who are trusted testers?
Google told us they are not the Search Quality Raters. Instead, they are a demographically and geographically diverse group of people external to Google.
These testers test Google’s products to help Google improve and understand how real users are likely to experience these products. Google said they work with a third-party supplier to provide this testing.
It is unclear how you or I can test out Bard before it launches more widely.
Attribution and citations
I asked Google how it plans to provide attribution and/or citations for its answers, either from Bard or in search results. Google did not have an immediate comment at the time this was published.
Why we care
It is fascinating to see how search and AI companies are adding human-like question-and-answer responses. With Bard, Google seems to be taking a similar approach to ChatGPT thus far, but it is early and I expect the user experience and interfaces will evolve rapidly.
How Google may handle attribution and citation – especially in search – is an interesting topic as well. Google has used AI to write knowledge panels for years, which is why Google does not always show attribution for some knowledge panels.
Earlier this week, we reported Google had started work on its own version of ChatGPT, named Apprentice Bard, as well as testing a new search design integrating chat:
When a question is entered, the search results show a gray bubble directly under the search bar, offering more human-like responses than typical search results. Directly beneath that, the page suggests several follow-up questions related to the first one. Under that, it shows typical search results, including links and headlines.
Here’s how Google responded to the report:
“We have long been focused on developing and deploying AI to improve people’s lives. We believe that AI is foundational and transformative technology that is incredibly useful for individuals, businesses and communities, and as our AI Principles outline, we need to consider the broader societal impacts these innovations can have. We continue to test our AI technology internally to make sure it’s helpful and safe, and we look forward to sharing more experiences externally soon.”
What about Microsoft and ChatGPT?
Meanwhile, Microsoft has invested heavily in OpenAI, the creator of ChatGPT. Microsoft is rumored to be bringing GPT-4 to Bing search in some way in the near future. And we have already seen early screenshots of Bing’s interface with ChatGPT-style integrated into search.
Many companies, especially those in search, are taking these technologies seriously after the explosive popularity of OpenAI’s ChatGPT.