Apple's Ferret: The New Frontier in Open Source AI that Surpasses GPT-4

Apple unveils Ferret, an open-source multimodal AI surpassing GPT-4. Capable of processing text, images, sounds, and videos, Ferret heralds major AI advancements.

Introduction:

The race for innovation in the field of artificial intelligence (AI) has just intensified with Apple's striking entry and its multimodal open source language model, Ferret. This significant development positions Ferret not only as a revolutionary tool in AI technology but also as a formidable competitor to the renowned GPT-4 by OpenAI.

The Emergence of Ferret, a Revolutionary AI by Apple

By unveiling Ferret, Apple boldly steps into the arena of artificial intelligence, marking a decisive turn in its technology strategy.

Spearheaded by Zhe Gan, a prominent AI researcher at Apple, and in partnership with experts from Columbia University, Ferret is the product of collaborative research and innovation efforts.

Multimodal Capabilities Surpassing Existing Models

Ferret stands out from its predecessors like Gemini, ChatGPT, and Google Bard by its ability to process not only text but also images, sounds, and videos.

This versatility allows it to tackle complex tasks with remarkable ease and efficiency, thus opening new avenues in AI applications.

Superiority in Image Analysis: A Giant Leap Beyond GPT-4

Ferret excels particularly in the field of image analysis, demonstrating a superior capability over GPT-4 in analyzing and interpreting specific image areas.

This performance is all the more remarkable as it comes with a significant reduction in errors, representing a major advancement in visual processing by AI.

Unmatched Computational Power Thanks to Nvidia Infrastructure

To achieve these levels of performance, Apple has equipped Ferret with eight Nvidia A100 GPUs, cutting-edge components in the field of AI.

These graphics processors, equipped with 80 GB of HBM2e RAM, are essential for managing the intensive computational load required by Ferret's training and operation.

Towards Integrating AI into Our Smartphones

Apple's ambition doesn't stop there: the company aims to make Ferret compatible with smartphones.

Although the challenges are considerable, especially due to the complexity of language models, recent advances suggest that the combined use of RAM and integrated flash memory could pave the way for advanced AI assistants on mobile devices.

Conclusion

With Ferret, Apple redefines the boundaries of what's possible in the field of generative and multimodal AI. By combining multimodal capabilities with a state-of-the-art computational infrastructure, Ferret not only positions itself as an important milestone for Apple but also as a cornerstone for the future of artificial intelligence.

As we eagerly anticipate the integration of this technology into our daily lives, Apple continues to chart a path of innovation and excellence.

TAGs
Articles récents