Google's Gemini beats OpenAI's ChatGPT by 3% in multi-discipline tests: Report

NEW DELHI, Feb 19 (IANS) – Google’s AI chatbot Gemini has surpassed OpenAI’s ChatGPT in a series of comprehensive multi-discipline tests, according to a recent report. The assessment, based on Stocklytics.com data, revealed that Gemini outperformed ChatGPT by three per cent. Google’s custom-designed tensor processing units and robust infrastructure were credited as driving factors behind Gemini’s superior performance, enabling efficient training and deployment for real-time applications.

Stocklytics Financial analyst Edith Reads highlighted the technical innovations powering Gemini, which extend beyond language processing to include remarkable performance in processing images, videos, and audio. The report also detailed Gemini Ultra’s superior performance in subjects such as Art and Design, Health and Medicine, Human and Social Sciences, and Technology and Engineering, outshining the GPT-4 in these areas. However, GPT-4 outperformed Gemini in Business and Science subjects.

The Gemini AI system includes three versions – Gemini Ultra, Gemini Pro, and Gemini Nano. Meanwhile, Google is expanding the availability of Gemini models on its Vertex artificial intelligence (AI) platform, with Gemini 1.0 Pro now generally available to all Vertex AI customers for scaling across AI tasks.

The report’s findings indicate Google’s continued innovation and investment in advancing AI technology, with Gemini emerging as a competitive chatbot in the multi-discipline testing landscape. As Google brings more Gemini models to customers and expands availability on its Vertex AI platform, the AI chatbot’s capabilities are expected to evolve and drive further advancements in real-time applications.

–IANS
shs/prw