The Google I/O 2024 has commenced with several announcement focused on advanced Gemini AI features, and now coming to several Google apps and services. Sundar Pichai during the keynote revealed advanced AI features for Google search, Workspace, Photos, and other apps. After making the Gemini 1.5 Pro available to all users, Google announced a new powerful and faster AI model. Demis Hassabis, CEO of Google DeepMind took the stage during the Google I/O where he revealed a new lighter-weight multimodal, “Gemini 1.5 Flash” which is cost-efficient, but faster. The new AI model comes with 1 million token context window and allows users to prompt with text, images, audio and video.
The Gemini 1.5 Flash is the new cost-effective AI model added to the Gemini family. Google highlighted that the AI model is capable of conducting high-volume, high-frequency tasks at a scale. Good said, “While it's a lighter weight model than 1.5 Pro, it's highly capable of multimodal reasoning across vast amounts of information and delivers impressive quality for its size.”
The Gemini 1.5 Flash is capable of conducting several tasks such as summarisation, chat applications, image and video captioning, data extraction from long documents and tables, and more. Google reported that the 1.5 Flash trained on a process called “distillation.” This process simply transfers knowledge and skills from a larger model to a smaller AI model, making it more efficient. The AI model is available for Google AI Studio, Vertex AI, Google Cloud, and developers.
On the contrary, the Gemini 1.5 Flash is a clear competitor for OpenAI's new GPT-4o model which was announced yesterday at the “Spring Updates” livestream. While the functionalities are quite different, the idea behind both models is very similar. Now, the question arises, which AI model is more powerful and capable of conducting complex day-to-day tasks? Apart from Gemini 1.5 Flash, Google also announced the Gemini 1.5 Pro for the public with a 2 million token
Read more on tech.hindustantimes.com