Google unveiled Gemini on Wednesday, which is claimed to be the most capable generative AI model till date. According to the company, this new model is more complex and can understand and respond to commands with a greater degree of nuance.
The company boasts of having produced a better, more sophisticated quality of technology with Gemini.
Sundar Pichai said, “”This new era of models represents one of the biggest science and engineering efforts we’ve undertaken as a company.”
Google says that the new AI is much more powerful than GPT-4, with five times the computational power. This means it can be trained faster and might be able to handle larger amounts of data.
According to Google, it is also the first model that can perform better than human experts on Massive Multitask Language Understanding (MMLU), a common way to test how well AI models know things and solve problems.
Starting from December 13, developers can use Gemini through Google Cloud’s API. There’s also a more advanced version of the AI expected in 2024, but it will only be released after thorough trust and safety checks.
Gemini has been trained to understand and reason with all kinds of input better than existing models.
Since Microsoft backed Open AI launched ChatGPT last year, Google has been racing to release its version of the AI.
Google is said to add its new AI technology to its AI assistant Bard and will launch a more powerful version of the model through Bard later next year.
Demis Hassabis, the founder of DeepMind, Google’s advanced AI lab, stated that has the ability to comprehend the world similar to humans. He emphasized that it surpasses all other existing models in terms of its capabilities.
Google DeepMind AI unit has made its largest AI model “Gemini” till date which they say is comparatively cheaper for allot to users in comparison with its previous models.
Gemini Models
Gemini Ultra is designed to be the most extensive and powerful model, specifically for intricate and demanding tasks. On the other hand, Gemini Pro excels in handling a broad spectrum of tasks, offering versatility. Meanwhile, Gemini Nano is optimized for on-device operations. Starting today, Gemini Nano is accessible in Pixel 8 Pro, enhancing features such as summarization in the Recorder app and Smart Reply in Gboard, beginning with WhatsApp. Google plans to introduce the new AI model across various products and services, including Search, Ads, Chrome, and Duet AI.
Comments 1