Google unveils Gemini, its AI model capable of competing with GPT 4. Google has taken the most significant step forward in its AI journey with the release of Gemini 1.0, its latest and greatest AI model.
Google unveils Gemini
Google Bard, the company’s generative AI chatbot and ChatGPT competitor, will now be driven by Gemini, which Google promises will dramatically improve its capabilities. Sundar Pichai, CEO of Google, termed it the “biggest science and engineering efforts we’ve undertaken as a company.”
Google describes Gemini as “natively multimodal,” because it was trained on images, video, and audio rather than just text.
Demos provided by Google
Google researchers also demonstrated how Gemini may aid scientific research by answering questions regarding a research report that included graphs and mathematics. Gemini could analyse a photo of a person’s arithmetic assignment and determine correct and incorrect answers.
Alphabet said it is developing three versions of Gemini, each with a different level of computing power. The most powerful version, Ultra, is intended for data centres. It will be available in early 2024. Pro is the medium-sized variant that is claimed to be the most capable all-purpose device and is now utilised for Bard. According to the business, the smallest, Nano, would run efficiently on mobile devices.
When will Google Gemini be available?
Gemini, Pro, and Nano were released in 170 countries on Wednesday. Google has made Gemini Nano available for the Pixel 8. The Pixel 8 Pro is the first phone to support Nano, allowing for new capabilities such as a “Summarise in the Recorder” app and Smart Reply for WhatsApp. In early 2024, Google will release “Bard Advanced,” which will feature Gemini Ultra, its most powerful iteration.
Which is Better, Gemini or ChatGPT?
According to Google, the “Ultra” version of Gemini can outperform ChatGPT. Gemini Ultra has already outperformed the OpenAI model in 30 of 32 areas, according to tests. The most proficient version of Gemini outperformed GPT-4 on multiple-choice examinations, grade-school math, and other benchmarks, according to a white paper released Wednesday, but highlighted persistent challenges in developing AI models to acquire higher-level reasoning skills.
Demis Hassabis, CEO and Co-Founder of Google DeepMind revealed in a blog post that Gemini beats humans in several linguistic tasks. He went on to say, “With a score of 90.0%, Gemini Ultra is the first model to outperform human experts on MMLU (massive multitask language understanding), which uses a combination of 57 subjects such as maths, physics, history, law, medicine and ethics for testing both world knowledge and problem-solving abilities.
