Alphabet Unveils Enhanced AI Search and Chatbot Amid Rising Competition

MOUNTAIN VIEW, California, May 14 (Reuters) - Alphabet, the parent company of Google (GOOGL.O), revealed on Tuesday its advancements in artificial intelligence across various sectors. This includes enhancements to its Gemini chatbot and upgrades to its renowned search engine. These developments highlight Google's response to heightened competition in the AI field, particularly following the introduction of ChatGPT by Microsoft-backed OpenAI in 2022.

Google recently introduced several innovations, including the Gemini 1.5 AI model Flash, aimed at enhancing speed and cost efficiency. Another project, Project Astra, enables real-time conversations based on smartphone camera captures. Additionally, AI-generated headlines categorize search results. Alphabet CEO Sundar Pichai views these updates as opportunities for growth, despite potential impacts on Google's profitability.

Google's product presentation at its annual I/O developer event in Mountain View, California, came shortly after a demonstration by rival OpenAI. OpenAI showcased ChatGPT's ability to respond with human-like intonation to various prompts. This highlights the intense rivalry between the two companies in the AI space, with both making significant strides in advancing AI technology.

For instance, Alphabet's AI unit, Google DeepMind, has been developing technology aimed at assisting consumers with daily tasks. One of its projects, Project Astra, utilizes smartphone cameras to interpret the surrounding environment. During a Google I/O demo, it identified a speaker and located misplaced glasses. The company also teased Gemini Live, offering a more natural voice and text assistant than its previous Google Assistant.

Google is also venturing into video generation with Veo, an AI model capable of creating 1080p-resolution videos exceeding a minute in duration. This offering, available to approved creators on a preview basis, includes filmmaker Donald Glover among its users. This move mirrors OpenAI's promotion of similar film-conjuring software in Hollywood, captivating and concerning the creative industry.

Additionally, Google announced enhancements to its Gemini Pro 1.5 model, which can interpret vast amounts of data. Doubling its capacity to 2 million tokens, this AI can process extensive text or video content, potentially answering questions based on thousands of pages of text or more than an hour of video material.

The Pro model, initially supporting prompt sizes of up to 1 million tokens, will be accessible to subscribers of Google's Gemini Advanced service. Alphabet's shares rose by 1% to $172.59 on Tuesday afternoon.

NEW CHIP, NEW SEARCH

Google also revealed its endeavors to enhance AI capabilities through new computing chips and updates to its search engine. It introduced a sixth-generation tensor processing unit (TPU) aimed at providing an alternative to Nvidia's powerful processors, available to Google Cloud customers by late 2024. Additionally, AI will organize search results for various queries, including dining and recipes, expanding to movies and books. The company introduced AI Overviews for complex queries and aims to integrate AI innovations into profitable products and services. Ads will remain on Google web pages, with AI Overviews expected to reach over a billion users by year-end.

Source: Reuters