New Delhi: Alphabet Inc’s Google is bringing more artificial intelligence to its core search product, hoping to generate the same consumer enthusiasm generated by Microsoft Corp’s updates to rival search engine Bing in recent months. At its annual I/O conference in Mountain View, California, on Wednesday, Google introduced a new version of its namesake engine.
Called Search Generative Experience, the new Google can generate answers to open-ended questions while maintaining its own recognizable list of links from across the web. “We’re reimagining all of our core products, including search,” Alphabet CEO Sundar Pichai said after taking the stage at the event. ,Also read: Explainer: How is the new Google AI search bard different from chatbots?,
He said Google is integrating generative AI into search as well as products like Gmail, which can create draft messages, and Google Photos, which can make changes to images like centering images and filling in blank space. ,Also Read: Google unveils Pixel 7a, foldable phone and AI features for search,
Vice President Cathy Edwards said in an interview that US consumers will gain access to the Search Generative Experience through a waiting list in the coming weeks, a testing phase during which Google will monitor the quality, speed and cost of search results.
After startup OpenAI introduced Silicon Valley’s darling chatbot, ChatGPT, which is better known as generative AI, Google started a furious funding race among competitors. Generative AI can create completely new content, such as completely composed text, images, and software code, using past data.
“AI can provide insight,” Edwards said. “But what people basically want at the end of the day is to be connected to information from real people and organizations, for example, knowing that this health information comes from WHO,” or the World Health Organization.
Explaining how AI can deliver misinformation, Edwards said the company prioritized accuracy and cited reliable sources. Google will also mark images generated with AI and make it easier for people to check the authenticity of the picture.
“Google’s approach makes a strong case that search is evolving, not dissolving, and Google is here to stay,” said Kingsley Crane, analyst at Canaccord Genuity.
What to wear?
Even with the AI embedded, Google still looks and acts like its familiar empty search bar.
But while a search for “weather San Francisco” would normally point a user to an eight-day forecast, a query asking what outfit to wear in the California city prompts a longer response generated by AI, According to an appearance earlier in the week for Reuters.
One challenge of drawing on such AI, known as large language models, is the high expense. “We and others are working on different ways to reduce costs over time,” Edwards said.
Ads will remain key, Edwards said. “We only pay for clicks.”
bard for 180 countries
“The company is showing a willingness and ability to reinvent and disrupt itself, which I think will be received favorably by investors,” said Michael Ashley Schulman, chief investment officer at Running Point Capital Advisors.
In February Google announced its competing chatbot called Bard. The company said on Wednesday that Bard will now be multimodal like OpenAI’s GPT-4, making the chatbot accessible to people in more than 180 countries and territories.
This means customers will be able to prompt Bard with images, not just text — for example asking the chatbot to write a caption to a picture they’ve been assigned, it said.
Behind Bard is also a more powerful AI model Google calls PaLM 2, which it said can solve tougher problems. Pichai also said that one of its PaLM 2 models is light enough to work on a smartphone.