To assist the development of generative AI applications, Amazon’s subsidiary Amazon Web Services (AWS) introduced Amazon Bedrock and other services on April 13. According to the company’s blog post, AWS is looking to leverage its existing machine learning capabilities and infrastructure to provide services to companies wanting to build generative AI applications. Generative AI tools or services are those AI systems which can create content, images, videos, music, etc. based on user prompts. Why it matters: A lot of the discussions concerning generative AI largely focus on the manner in which Big Tech is using data to train these tools. In this case too, whose data is being used for training AWS machine learning models is something worth exploring. When it comes to tackling harms caused by generative AI, experts say a lot depends on how these systems are trained and what the approach of the company developing them is; for example, Mozilla recently introduced Mozilla.ai to in assist building “Trustworthy Ai products”. Keeping track of how Big Tech companies like Google, Amazon and Microsoft are experimenting with generative AI is important to evaluate industry responses to the debates about accountable and responsible AI. What’s AWS offering? 1. Amazon Bedrock: The new service is expected to allow AWS users access to foundation models (FM) or pre-trained ultra-large machine learning (ML) models used for “building and scaling” generative AI applications. According to a blog post by AWS, customers will have access to FMs from top AI startup model providers such as AI21,…
