AI search storm
AI search storm
Microsoft's Bing Generative Search
Google AI Overviews went from 84% to 15% of searches
***
AI search storm - AI's search quake shakes media landscape
The launch of new AI search products, including OpenAI's SearchGPT and Microsoft's Bing Generative Search, is prompting news organizations to reevaluate their potential collaborations with technology companies that rely on news content for real-time event queries.
Historically, negotiations between the tech and the news sector have centered on supplying data for the extensive training of large language models (LLMs). However, recent discussions are shifting focus toward more specific use cases, where news publishers may find themselves in a stronger negotiating position. While LLMs can be trained on vast amounts of text, they require access to smaller, curated datasets of verified information to effectively respond to current event inquiries.
The recent rollout of OpenAI's SearchGPT and Microsoft's Bing Generative Search has shed light on the evolving nature of partnerships between major tech firms and news publishers, as LLM developers increasingly incorporate fact-based news methodologies into their offerings.
Despite these advancements, it remains uncertain whether the new generative AI-powered search engines will generate revenue for publishers comparable to that of traditional search platforms like Google. A spokesperson indicated that the agreements with news publishers for fact-checked content are structured around licensing fees rather than revenue-sharing models….
Originally by Sara Fischer, Scott Rosenberg of Axios
***
Microsoft's Bing Generative Search
“After introducing LLM-powered chat answers on Bing in February of last year, we’ve been hard at work on the ongoing revolution of search. Today, we’re excited to share an early view of our new generative search experience which is currently shipping to a small percentage of user queries.
By combining the power of generative AI and large language models (LLMs) with the search results page, Bing’s Generative Search creates a bespoke and dynamic response to a user’s query. For example, if a user searches "What is a spaghetti western?" Bing shows an AI-generated experience that dives into the film subgenre, including its history and origins, top examples and more. The information is easy to read and understand, with links and sources that show where it came from or let the user dive deeper. The regular search results continue to be prominently displayed on the page like always.
This new experience combines the foundation of Bing’s search results with the power of large and small language models (LLMs and SLMs). It understands the search query, reviews millions of sources of information, dynamically matches content, and generates search results in a new AI-generated layout to fulfill the intent of the user’s query more effectively….
Originally by Microsoft Bing Blogs
***
Google AI Overviews went from 84% to 15% of searches
BrightEdge, the global leader in AI-driven organic search, content, and digital marketing automation, has released new data from the BrightEdge Generative Parser™. Major findings include a steep reduction in responses containing AI Overviews, Featured Snippets serving as a strong predictor for generative AI responses, and improvements in Google’s ability to provide precise responses and anticipate follow-up questions. Following Google's official release of AI Overviews (previously called Search Generative Experience).
Jim Yu, Founder and Executive Chairman at BrightEdge. "In the past, search engines would pull a lot of information and give you links. Now, AI helps do the searching, summarizes the results, and provides an informative opinion.
Google has gradually been reducing AI in search results from 84% to under 15%. Historically, AI was present 84% of the time in opt-in or collapsed states. Since mid-April, Google has slowly been removing Opt-in results to the point where they now barely show up. Today, AI Overviews show up mainly as collapsed in less than 15% of search results. The likely reason is to reduce the risks of incorrect AI answers while refinements are made in a public environment.
Originally By Jim Yu, Founder and Executive Chairman at BrightEdge