Google began Real-time search functions for its Gemini AI platform on Thursday, allowing its language models to access up-to-date information from Google Search. The recent feature called “Grounding with Google Search“” is aimed toward developers constructing AI applications and is different from OpenAI’s consumer-focused ChatGPT search service, which launched on the identical day.
“We’re focused on integrating search-powered answers into developer workflows,” said Logan Kilpatrick, head of product at Google, in an exclusive interview with VentureBeat. “We’re making the most of what Google does uniquely well: making the world’s information accessible through search.”
Say hello to Grounding with Google Search, available in Gemini API + Google AI Studio!
You can now access real-time, up-to-date information from Google Search when constructing with Gemini by enabling the grounding tool.https://t.co/oGVTOKHfM8
— Logan Kilpatrick (@OfficialLoganK) October 31, 2024
The system allows developers to complement their AI applications with recent search data, complete with citations and sources. The service costs $35 per 1,000 queriesreflecting the numerous computational needs for real-time AI search.
The technology uses a “dynamic retrieval” system that robotically determines when to access search results. Each query receives a rating between 0 and 1 – current events questions rating high (0.97), while creative writing prompts rating low (0.13). This helps manage each costs and response times while ensuring accuracy.
In the $49 billion battle for the long run of search
Google's move to integrate search into its AI platform comes at a critical time. The company deserves it $49.4 billion from search promoting within the third quarter of 2024but is facing growing pressure from AI-supported alternatives. Running these systems requires massive computing resources – OpenAI estimates $5 billion shall be spent on computing costs this 12 months alone.
The integration also raises questions on publisher compensation. Both Google and OpenAI have licensing deals with major news organizations, however the financial terms remain confidential. Several publishers, including The New York Timeshave filed lawsuits against AI systems that use their content without permission.
Why OpenAI's recent ChatGPT search could change the best way we discover information online
Hours after Google's announcement, OpenAI launched ChatGPT Search and took a special approach by targeting consumers directly. While Google focuses on providing tools for developers to construct search-enhanced AI applications, OpenAI's service offers end users the power to access up-to-date information on news, sports, stocks and weather through a conversational interface – notably without promoting.
“The journey we're on is to make use of Google Search in additional creative ways and across multiple interfaces,” said Shrestha Basu Mallick, Google's group product manager for the Gemini API, in an interview with VentureBeat. “You get it through AI Studio, the Gemini APIs, and it may eventually turn into native to the model itself.”
This recent phase of competition could change the best way people find information online. Instead of scrolling through pages of results, users may increasingly depend on AI systems to synthesize answers from multiple sources. But questions remain about accuracy, publisher compensation, and whether firms can construct sustainable business models around these compute-intensive services.
The simultaneous launches suggest that AI-powered search could turn into a three-way battle between Google, Microsoft (through its OpenAI partnership) and OpenAI itself.
Google continues to have benefits in search infrastructure and promoting revenue, while OpenAI has demonstrated its capabilities in developing compelling AI products for consumers. Microsoft, meanwhile, advantages from each through its billion-dollar OpenAI investment.
