Google introduces new features in AI combat with ChatGPT

Google has announced a new artificial intelligence model designed to challenge ChatGPT as the online AI arms race escalates.

EO Sundar Pichai says his new conversational AI service called Bard will soon be available for the public to experiment and play with. It will initially be opened to a smaller number of testers. The company is also introducing some new AI features in search.

It could mean that the way we use search engines every day has fundamentally changed, as more “natural” questions lead to more accurate and complete answers.

Google is the dominant search engine in Europe and is used for over 90 percent of search engine queries.

But since opening to the public late last year, ChatGPT has provided a powerful new glimpse into what might be possible for search in the future, allowing public users to search essays, art, or computer code in seconds in response to boxed requests to get a specific style or format. The system developed by OpenAI is so compelling that it has sparked debate in schools and colleges about whether students can cheat on homework by submitting flawless essays “written” by ChatGPT.

Since its launch, ChatGPT has been tested by hundreds of millions of people and has now attracted a multi-billion dollar investment from one of Google’s biggest competitors, Microsoft.

“We’re also sharing a look at several AI capabilities you’ll soon see in search that can help synthesize insights for questions that don’t have a right answer,” Mr. Pichai said in a blog post Monday.

Google, he said, “is also working on new generative language APIs, starting with LaMDA, to equip developers, makers, and enterprises to develop their own AI products in the future.”

Some technologists have warned about the power of systems like ChatGPT and other advanced conversational AI tools to generate and share misinformation or be used for other destructive purposes.

Mr. Pichai said the company is aware of the power of such systems.

“Our approach to AI has always been bold, but we also use it responsibly,” he said.

“Generative AI is playing an increasingly important role in society, so it is imperative that we consider the broader implications of our technology. That’s why we’re committed to introducing these features responsibly, to ensure we meet our bar for safety, quality and down-to-earthness, before making them more widely available to the public. It’s a really important balance to get it right.”