“Two years ago we unveiled next-generation language and conversation capabilities powered by our Language Model for Dialogue Applications (or LaMDA for short),” CEO Sundar Pichai said in a blog post. “We’ve been working on an experimental conversational AI service, powered by LaMDA, that we’re calling Bard.”
Pichai said Bard processes information from the internet to generate output responses for input user queries. Bard can, for example, do things like explain new discoveries from NASA’s James Webb Space Telescope to children, provide up-to-date information on the best strikers in football and come up with training exercises to help people improve their health.
Bard looks like it rivals ChatGPT’s ability to be conversational but appears to be more focused on search and explainability rather than being a general language model capable of writing poems or essays. A preview snippet of Bard in action shows the model generates responses in bullet points.
“When people think of Google, they often think of turning to us for quick factual answers, like ‘how many keys does a piano have?’ But increasingly, people are turning to Google for deeper insights and understanding — like, “is the piano or guitar easier to learn, and how much practice does each need?” Learning about a topic like this can take a lot of effort to figure out what you really need to know, and people often want to explore a diverse range of opinions or perspectives,” Pichai explained.
Bard will be launched to the public in the coming weeks, and will initially be powered with a smaller model version of LaMDA. Google has released it to a select group of users for testing, and will be assessing the model’s safety and quality of responses.
Source: Google unleashes fightback to ChatGPT, Bard • The Register