Google has unveiled Gemini, a groundbreaking AI model designed to transcend traditional boundaries by seamlessly interpreting text, images, audio, video, and code simultaneously. Aptly named, Gemini boasts three distinct sizes—Ultra, Pro, and Nano, each tailored for specific tasks, from highly complex endeavours to on-device efficiency.
Gemini Ultra emerges as the pinnacle, unravelling the intricacies of Python, Java, C++, and Go code, excelling in benchmarks such as HumanEval and Natural2Code. Its prowess extends beyond coding, demonstrating exceptional aptitude in explaining intricate subjects like math and physics.
The Pro variant, harnessed by Google’s Bard AI, promises heightened reasoning, planning, and understanding, potentially challenging the likes of ChatGPT. According to Google, Gemini Pro outshone GPT-3.5 in six out of eight benchmarks, solidifying Bard’s position as the preferred free chatbot.
Surprisingly, the Pixel 8 Pro takes centre stage as the first smartphone engineered for Gemini Nano, unleashing its potential through features like Summarize in Recorder and Smart Reply in Gboard. While Bard with Gemini Pro is available for English text prompts in numerous territories, the enhanced Gemini Ultra experience is slated for a release early next year.
Google’s experimentation with Gemini extends to Search, promising accelerated experiences. In the pipeline are plans to integrate Gemini into various Google products and services, from Ads and Chrome to Duet AI. The company’s visionary approach envisions Gemini as a transformative force, shaping the future of AI across diverse applications and platforms.