Gemini vs. ChatGPT: Google’s AI Dominates in 30 Out of 32 Test Categories

A new Pro model of Google’s most recent artificial intelligence, Gemini, has been introduced. According to sources within the firm, it has outperformed the free version of ChalGPT in extensive testing. According to performance reports, Gemini Ultra outperforms the present state-of-the-art results on thirty of the thirty-two academic benchmarks. These benchmarks are frequently utilized in the study and development of large language models (LLM).

Google has been accused of falling behind OpenAI’s ChatGPT. It is widely considered to be the most capable and popular artificial intelligence platform currently available. Gemini was reportedly trained to be multimodal, which means that it is capable of processing many sorts of media like text, photos, video, and audio, according to Google.

According to the report, Gemini Ultra has achieved a score of 90.0% on MMLU surpassing human experts in this domain. MMLU tests knowledge of 57 subjects, including math, physics, history, law, medicine, and ethics, as well as problem-solving skills.

Google Unveils Gemini: AI’s New Era with Advanced Reasoning

The Google-powered AI, Gemini, will be available in three variants: Ultra, the flagship model; Pro; and Nano (for smartphones). The company claims that on December 13, Gemini Pro will be available to developers via AI Studio and to enterprise clients via its Vertex AI program. It is also reportedly possible to get the Pro version through Bard, the company’s chatbot interface.

According to Eli Collins, Vice President of AI Products at DeepMind, the Ultra version is capable of understanding “nuanced” information found in text, images, audio, and code. Collins revealed that some of the information utilized in the app’s development came from openly accessible websites. However, he did not specifically specify the sources used to train the AI.

Check out: Google Pushes Gemini AI Launch to Early 2024.