Google Just Launched Gemini, Its Long-Awaited Answer to ChatGPT
Google has just unleashed its highly anticipated response to ChatGPT, and it’s called Gemini. In the ever-evolving landscape of artificial intelligence, Google aims to reclaim its top spot with this groundbreaking AI project. One year after the debut of OpenAI’s ChatGPT, the race to advance AI technology has only intensified.
Gemini is not just another AI model; it’s a game-changer. Unlike its predecessors, Gemini is a truly multimodal AI, handling text, images, and video seamlessly. This could mark a pivotal moment in Google’s history, akin to the impact PageRank had on transforming the search engine into a global phenomenon.
The initial rollout of Gemini is happening within Google’s Bard chatbot, tailored for English language settings and accessible in over 170 countries. Developers, brace yourselves – Gemini will be up for grabs via Google Cloud’s API starting December 13. If you’re a Pixel 8 smartphone user, you’re in for a treat as a streamlined version of Gemini powers suggested messaging replies right from your keyboard.
Demis Hassabis, Google DeepMind’s CEO, expressed enthusiasm, calling it a significant moment for the company. The potential of Gemini has sparked excitement not just within Google but also among developers eager to explore its capabilities.
Gemini boasts three versions: Ultra, the most powerful; Nano, a more efficient but smaller variant; and Pro, striking a balance between size and capability. The rollout includes Gemini Pro in Google’s Bard, enhancing its reasoning and planning abilities. This week, a specialized Gemini Pro version joins a revamped AlphaCode, a coding generative tool from Google DeepMind. The ultimate version, Gemini Ultra, is slated for a 2024 release, pending rigorous trust and safety evaluations.
Sissie Hsiao, Google’s Vice President and Bard’s General Manager, highlighted how Gemini’s multimodal prowess has elevated Bard’s skills. Tasks like content summarization, brainstorming, writing, and planning have seen significant improvements, making Bard more versatile than ever.
Gemini’s potential extends beyond traditional AI applications. Demonstrations showcased its ability to tackle visual challenges, responding adeptly to videos featuring drawings, puzzles, and requests for game ideas. In scientific research, Gemini proved its worth by answering questions related to research papers, complete with graphs and equations.
Eli Collins, Google DeepMind’s Vice President of Product, revealed that Gemini Pro outperformed the earlier ChatGPT model, GPT-3.5, on six out of eight commonly used AI benchmarks. Gemini Ultra, set to debut in 2024, claims an impressive 90% score on the Massive Multitask Language Understanding (MMLU) benchmark, outperforming even GPT-4.
Google confidently asserts Gemini’s state-of-the-art performance across a wide range of benchmarks, setting new standards in the machine-learning research community. As the generative AI boom gains momentum, Gemini emerges as a frontrunner, signaling a paradigm shift in the world of artificial intelligence. The question now is, what groundbreaking applications will developers unlock with Gemini’s capabilities? Only time will tell.
Yep, I went through some videos and articles to get to know the Gemini!
As per my study I can understand that Gemini represents a colossal scientific and engineering endeavor, a testament to the collaborative efforts of teams across the company, including Google DeepMind and GoogleResearch. Heartfelt appreciation goes out to all the dedicated Googlers whose contributions made Gemini a reality!