As OpenAI continues to dominate the headlines, Microsoft and Amazon have responded with their own groundbreaking AI innovations, showcasing their ambitions to rival OpenAI in assistive and generative AI.
Microsoft’s Copilot Vision
On 5th December, Microsoft announced Copilot Vision, a feature integrated with its Edge browser, allowing users to interact directly with web content. Available through the Copilot Labs preview, the tool answers questions, summarises, translates, and highlights content on web pages. Users can, for example, find recipes or gaming tips with simple queries.
To address privacy concerns, Microsoft emphasises that Copilot Vision processes and deletes session data without storing it for model training. Initially limited to pre-approved websites, the tool avoids paywalled and sensitive content, a cautious approach considering legal challenges from publishers like The New York Times.
Amazon’s Nova AI Model
Just a day later, on 6th December, Amazon launched Nova AI models, which is their powerful alternative in the generative AI arena. These models, accessible via AWS Bedrock, are designed for tasks ranging from text and image creation to complex reasoning and video generation. With costs up to 75% lower than comparable tools, Amazon aims to attract enterprises looking for customisable AI solutions.
Key Nova models include:
- Nova Micro for fast text processing.
- Nova Canvas for image generation.
- Nova Reel for video creation.
Nova models support over 200 languages and feature customisation capabilities, allowing businesses to fine-tune the AI using proprietary data. Amazon emphasises safety and responsible usage through detailed guidelines.
Competing for AI Supremacy
In this latest battle of the tech titans, both Microsoft and Amazon will aim to address the gaps left by OpenAI’s generative AI dominance. While Microsoft targets individual users with seamless web assistance, Amazon focuses on enterprise-grade solutions that promise affordability, scalability, and broad functionality.
