Innovating Google Search: Leveraging Advanced Generative AI Technology

2023-06-21

"In 2002, Larry Page once said, 'Google's mission can only be achieved when its search engine reaches a state of AI completeness.' Over twenty years later, this statement seems almost prophetic. The tech giant has long dominated the search field, with a market share of over 93%, while its biggest competitor, Microsoft, only holds a 3% share. However, in the power struggle between major tech companies, especially after the widespread adoption of OpenAI's ChatGPT supported by Microsoft, Google seems to be gradually losing its vitality. But now, Google is ready to take over. The company seems to have found its advantage through the introduction of a new AI-driven search experience at the 2023 Google I/O conference. Although it is not yet available to the public, several test users in the United States have already praised these new upgrades on social media. Suppressing Online Searches The Search, built on LLM PaLM 2, combines the capabilities of generative AI, allowing users to quickly understand complex topics through comprehensive answers. To facilitate further exploration, related topics are conveniently presented alongside search results. The new Search also features a ChatGPT-style interface, allowing users to ask follow-up questions and maintain context while delving deeper into the subject. Compared to ChatGPT, Bard performs better in research. Bard is also proficient in understanding Python backend to handle complex queries without explicit prompts from users, preventing it from generating illusions. Google has adopted this approach to enhance Bard's capabilities, surpassing System 1 thinking and incorporating elements of System 2 thinking. By integrating code into equations, a characteristic of System 2 thinking, Google aims to combine the strengths of both approaches for optimal performance. According to test user Linas Buenas, the improvements to Google's search engine now provide the ability to optimize search queries within the search results page, eliminating the need to start a new search if the initial results are not satisfactory. Additionally, users can perform various tasks without leaving the search page, simplifying the online experience and improving efficiency. One notable feature is that users can easily solve specific problems by copying and pasting code to receive assistance from Google. As expected, Google plans to integrate "written answers" similar to ChatGPT into its new AI search engine, which will be showcased alongside prominent domains. This emphasizes the importance of high-quality content in improving website visibility in search results. In summary, Google's focus is shifting from merely collecting information to actively editing and curating information. Seamless Online Shopping Google has introduced two new AI-powered features for virtual try-on, making it more diverse and inclusive. Based on a new diffusion architecture, "TryOnDiffusion: Two UNets' Story" showcases a variety of clothing on real models with different sizes, skin tones, and races. This update aims to make the virtual try-on experience more inclusive and representative, featuring real models ranging from XXS to 4XL. The AI model accurately demonstrates how clothing drapes, folds, stretches, and forms wrinkles and shadows on virtual models. However, these features are currently only available in the United States. The second feature uses machine learning and visual matching to optimize product recommendations. It helps shoppers find the perfect item by fine-tuning search criteria for color, style, and pattern. Unlike physical store shopping, it offers a wide selection from various online stores. Currently applicable to tops only, it improves the online shopping experience by allowing users to evaluate clothing and providing personalized search options before making a purchase. With Google's shopping graph, it may expand to more brands and products in the future. This effortless and enjoyable shopping experience is supported by Google's shopping knowledge graph, which contains over 350 billion product listings, suppliers, trademarks, and available inventory, compared to Amazon's less than 350 million products. Additionally, Google Lens helps you shop by simply taking a photo or screenshot to find and purchase your desired products. Furthermore, Google Lens can now identify skin conditions by taking photos or uploading images. This feature can also be used to identify other body issues, such as lip swellings, nail lines, or hair loss, making it easier to find relevant information. During travels, Lens becomes a valuable language translator, converting street signs, menus, and other texts into over 100 languages. For students, Lens provides step-by-step assistance with math, history, and science problems, supporting search results in multiple languages. Bold Advertising Bets According to the latest information, search advertising has become Google's main source of revenue in 2022. Out of the staggering $279.81 billion accumulated by the company, an astonishing $162.45 billion comes from search advertising. Therefore, Google's goal of profiting from AI-supported search is not surprising. Google and YouTube have launched two AI-based advertising solutions: Demand Gen and Video View. Demand Gen integrates efficient videos and images into platforms such as YouTube, YouTube Shorts, Discover, and Gmail. It provides advertisers with a simple building process to import successful assets into Google ads and preview the ad's effectiveness. Advertisers can improve their campaigns by customizing creatives and audience segmentation. Video View maximizes view counts in streaming ads, feed ads, and YouTube Shorts, surpassing traditional pay-per-view advertising campaigns. Following Amazon's lead, Google also prominently displays sponsored products. It showcases AI-generated content at the top of search results, followed by sponsored and non-paid posts. Previously, Google believed in ranking content based on expertise, but now it has shifted towards using AI-generated answers. This change suggests that Google may be exploring ways to generate additional revenue by allowing websites to pay Google when their content is selected to generate answers. Therefore, welcome to Google 2.0. It will only become bigger and better." Source: https://analyticsindiamag.com/googles-search-supremacy-reinforced-with-generative-ai/

RECENT AI NEWS

RECENT AI TOOLS