AWS CEO bats for responsible AI for unleashing GenAI's potential, managing risks
At the technology giant's annual flagship event here, Selipsky spoke about AWS and chip major NVIDIA's strategic collaboration to offer new supercomputing infrastructure software and services for Generative AI.
Advocating responsible artificial intelligence (AI), (AWS) CEO Adam Selipsky stressed the need for unlocking Generative AI's full potential while mitigating risks, and called for unprecedented collaborations and multi-stakeholder efforts spanning tech companies, policymakers, community groups and academia towards this.
Selipsky was speaking at the keynote of AWS re:Invent 2023, where he made a slew of announcements around strengthening Generative AI offerings. He affirmed that AWS is ready to help businesses reinvent with this "promising tool for innovation".
At the technology giant's annual flagship event, Selipsky spoke about AWS and chip major 's strategic collaboration to offer new supercomputing infrastructure software and services for Generative AI, and announced Amazon Q, a new generative AI-powered assistant specifically designed for work.
The company also announced the next generation of two AWS-designed chip families AWS Graviton4 and AWS Trainium2 that deliver advancements in price performance and energy efficiency for a broad range of customer workloads, including machine learning (ML) training and generative artificial intelligence (AI) applications.
The top honcho of Amazon's cloud computing unit said Generative AI is the next step in Artificial Intelligence, and will reinvent every application.
Selipsky also drew attention to the issue of responsible AI and highlighted the need to "find ways to unlock generative AI's full potential while mitigating risks".
"...the capabilities that make Generative AI such a promising tool for innovation possibly increase potential for misuse," he said.
Dealing with the challenge will require unprecedented collaborations and truly multi-stakeholder efforts across technology companies, policymakers, community groups, scientific communities and academia, he asserted.
"We have been actively participating in a lot of groups that have come together to discuss these issues," he said.
Earlier this year, the Biden administration in the US secured voluntary commitments from leading artificial intelligence companies to manage the risks posed by the new technology, and to help move toward safe, secure, and transparent development of AI.
"And just this month I joined UK PM (Rishi Sunak) for his AI Safety Summit to discuss new approaches including the formation of new AI safety institute," Selipsky said.
An important component of responsible AI is promoting interactions between consumers and applications that are safe, avoid harmful outputs, and stay within the company's guidelines, he said, adding that the easiest way to do it is to place limits on what information models can or can't return.
"Today we are announcing guardrails for Amazon Bedrock. Now this is a new capability that helps you (customers) easily safeguard your Generative AI applications with your responsible AI policies, to create a guardrail...," he said.
With this, customers can create multiple guardrails tailored to different use cases and apply them across multiple FMs, providing a consistent user experience and standardising safety controls across Generative AI applications, according to AWS.
Further, Selipsky noted that the innovation around Generative AI model is "explosive".
"GenAI will reinvent every application we interact with, at work and at home...From startups to enterprises, organisations of all sizes are getting going with Generative AI. They want to take the momentum that they are building with early experimentation, and turn it into real-world productivity gains," he said.
AWS is excited about the Large Language Model and other Foundation Models (FM) it is building.
Amazon Q, a new generative AI-powered assistant announced by AWS on Tuesday, is specifically designed for work and can be tailored to businesses to have conversations, solve problems, generate content, and take actions using the data and expertise found in the company's information repositories, code, and enterprise systems.
Amazon Q has a broad base of general knowledge and domain-specific expertise and is designed to help customers meet stringent enterprise requirements, says AWS.