Google Just Included Generative AI to Search

Image credit : NPR Google didn't seem in the least bit concerned about its search business SIX MONTHS AGO. Then Microsoft's Bing sprouted a chatbot, and OpenAI's ChatGPT was released. The search engine giant revealed today at its annual I/O conference that it will leverage generative artificial intelligence technology, like ChatGPT, to enhance results. The business is releasing a test version of their renowned search engine that includes text generation similar to that used by ChatGPT and other cutting-edge chatbots. Google's redesigned search still requires users to enter a query, and it still returns connections to other websites, content snippets, and advertisements as results. However, in some cases, the text at the top of the page will contain links to content that was created by AI using information gathered from various websites. In order to obtain more detailed information, a user can pose further queries. [caption id="attachment_171144" align="aligncenter" width="2500"] Image credit : Google Cloud Blog[/caption] A few paragraphs summarizing the coronation of Britain's new king could answer a question concerning it. In response to a question regarding ebikes, Google's algorithms can provide a bulleted list of product reviews from various websites along with links to online shops where a user can make a purchase. A new tool called Search Labs will make the updated version of search available to users in the US, but it won't be turned on by default or for everyone who uses Google. Compared to ChatGPT, Google's AI-powered search is far more subdued, forgoing an anthropomorphized avatar and steering clear of potentially divisive subjects like politics and expert medical or financial advice. For instance, Google's generative AI product declined to respond to queries about Joe Biden's suitability for the office of president or details regarding the abortion laws of various US states. Although Google is moving swiftly to include ChatGPT-like features into search, it is still unclear whether users will find them useful. For example, product searches combined information from several evaluations, but it wasn't immediately clear how the succinct summary may enhance the search process. [caption id="attachment_171147" align="aligncenter" width="1600"] Image credit : Search Engine Journal[/caption] These new features' unrefined appearance may be a sign that their introduction is a defensive measure. Over the past few years, Google has heavily invested in AI, and CEO Sundar Pichai has frequently referred to the firm as "AI first." However, the introduction of ChatGPT, an OpenAI chatbot that is remarkably witty and chatty but also fundamentally defective, caught Google off guard. A machine learning model that was trained to predict the words most likely to follow a string of text by consuming enormous volumes of text, including a sizable number of web pages, powers ChatGPT. ChatGPT underwent additional training to improve its ability to respond to queries and carry on conversations, which was provided by humans who graded the quality of the bot's responses. Even though ChatGPT is prone to faking information, users immediately discovered it to be a promising new way to search because it was trained on a large portion of the web. By investing $10 billion in OpenAI in January and integrating ChatGPT into Bing a month later, Microsoft seized this opportunity. Google was forced to scramble to catch up due to the buzz and intense user interest, which resulted in millions of users feeding inquiries to ChatGPT and Bing chat. Some of the core technology used in the new chatbots was created by corporate researchers, but Google was hesitant to publicly introduce LaMDA, the precursor to ChatGPT. [caption id="attachment_171148" align="aligncenter" width="1200"] Image credit : VOI[/caption] Google shifted course and unveiled Bard as a ChatGPT rival in March. Google said in April that it would integrate its AI research team with DeepMind, another Alphabet firm that specialises in AI. The technology for text generation is now being incorporated into Google's search engine. The power and utility of ChatGPT-style technologies may be put to the biggest test yet thanks to Google's prominence and reputation. It's also dangerous. Also read : Apparently, Elon Musk is launching a generative artificial intelligence project at Twitter Companies must create methods to verify that the information provided to users is accurate because language models are susceptible to fabrication. Additionally, some web publishers worry that search engines would reuse and scrape their content, which will result in fewer referrals. A substantial portion of Google's revenue comes from its search advertising business, thus the corporation must avoid cannibalizing that division.