(Credit: Tada Images/Shutterstock.com)
Amazon is rolling out new features for developers and businesses using its Q AI model.
Q can now write and debug code or make multi-line code suggestions. The tech giant is also offering a new Amazon Q Apps tool that generates custom-built AI-powered apps based on text prompts.
Q can also answer questions for employees at work about their company's operations. These custom chatbot features were first announced in November, but Amazon is now fully launching Q with its chatbot and coding capabilities to all its business clients. Q can also output data-driven narratives about a company's performance over time, diagnose AWS issues, or help IT pros improve a company's cloud services, for example. It can also help developers upgrade their software and analyze existing code.
Doug Seven, AWS's director and general manager of AI developer experiences, tells PCMag in an interview that Q can now plug into 40 different existing "connectors" or apps like Slack, Gmail, or Microsoft Exchange. This means AWS clients with Q could toss the AI into their company's Slack server and staff could ask the bot questions by mentioning it by name. Seven believes Q's potential is ultimately limitless for businesses, and Amazon is open to adding new plugins for Q over time based on what its users want (it already has firms like Toyota and Sun Life as clients).
"I have a sense of fatherly pride in all of this," Seven says, adding that Q can do much more now than when it was first announced last year, ultimately functioning as a main hub for users with AWS accounts. He shares that Q can write about 50 lines of code in just a few seconds and the AI was able to upgrade 1,000 older Java applications in a few days, for example.
“Since we announced the service at re:Invent, we have been amazed at the productivity gains developers and business users have seen," AWS VP of AI and Data Swami Sivasubramanian said in a statement. "Early indications signal Amazon Q could help our customers’ employees become more than 80% more productive at their jobs; and with the new features we’re planning on introducing in the future, we think this will only continue to grow."
While other AI coding tools like Microsoft's GitHub Copilot and the so-called AI software engineer Devin already exist, Q Apps shares some similarities with Devin in that Q Apps can also create AI-powered software projects from start to finish with just a text prompt. Amazon hopes people use Q Apps at work to get tedious tasks done more quickly. Employees could use Q Apps to streamline internal hiring processes, for example, Seven says, but he doesn't envision the app-generating tool being used in any consumer-facing capacity.
AI tools could speed up some mundane tasks at work, but they aren't always perfect. Last year, some Amazon employees reported Q was prone to "severe hallucinations and leaking confidential data." In response, Amazon said at the time that employees were sharing feedback internally, but claimed that no security issues were found.
Seven tells PCMag that Amazon has worked internally to validate Q's outputs and ensure they're correct. He also says Q hasn't given employees at any company information they weren't already allowed to have access to. "There's a lot of techniques we use in that infer pipeline to validate that code," Seven says. "We have a high level of confidence on it."
Editors' Note: This story has been updated to include comments from Doug Seven.