Pour one out for CodeWhisperer, Amazon’s AI-powered assistive coding tool. From today on, it’s kaput – sort of.
CodeWhisperer is now Q Developer, part of Amazon’s Q family of business-oriented generative AI chatbots that also extends to the newly announced Q Business. Available through AWS, Q Developer help with some of the tasks that developers perform during their daily work, like debugging and upgrading applications, troubleshooting, and running security scans, much like CodeWhisperer did.
In an interview with TechCrunch, Doug Seven, general manager and director of AI developer experiences at AWS, suggested that CodeWhisperer was a bit of a brand failure. Third-party metrics reflect as much; Even with a free tier, CodeWhisperer has struggled to keep up with the momentum of its main GitHub rival Copilot, which has more than 1.8 million paying individual users and tens of thousands of enterprise customers. (Bad first impressions sure didn’t help.)
“CodeWhisperer is where we started (with code generation), bBut we really wanted to have a brand – and a name – that fit a broader set of use cases,” Seven said. “You can think of Q Developer as the evolution of CodeWhisperer into something much broader.
To this end, Q Developer can generate code, including SQL, a programming language commonly used to create and manage databases, as well as test that code and help transform and implement new code created from database queries. developers.
Similar to Copilot, customers can fine-tune Q Developer on their internal codebases to improve the relevance of the tool’s programming recommendations. (CodeWhisperer, now deprecated, also offered this option.) And, thanks to a feature called Agents, Q Developer can autonomously perform tasks like implementing features and documenting and refactoring (i.e. say restructuring) of the code.
Ask Q Developer a query like “create an ‘Add to Favorites’ button in my app” and Q Developer will analyze the app code, generate new code if necessary, create a step-by-step plan, and perform performance testing. the application. code before executing the proposed changes. Developers can review and iterate on the plan before Q implements it, linking the steps together and applying updates to the necessary files, code blocks, and test suites.
“What happens behind the scenes is that Q Developer sets up a development environment to work on the code,” Seven said. “So in the case of feature development, Q Developer takes the entire code repository, creates a branch of that repository, scans the repository, does the work it’s been asked to do, and sends those code changes back to the developer. “
Image credits: Amazon
Agents can also automate and manage code upgrade processes, Amazon says, with Java conversions in effect today (specifically Java 8 and 11 built using Apache Maven to Java version 17) and upcoming .NET conversions. “Q Developer analyzes the code – looking for anything that needs to be upgraded – and makes all those changes before sending it back to the developer for review and commitment,” Seven added.
To me, Agents looks a lot like GitHub’s Copilot workspace, which similarly generates and implements bug fix plans and new features in software. And – as with Workspace – I’m not entirely convinced that this more autonomous approach can solve the problems with AI-based coding assistants.
An analysis of more than 150 million lines of code committed to project repositories over the past several years by GitClear found that Copilot was causing more bad code to be pushed to codebases. Elsewhere, security researchers have warned that Copilot and similar tools can amplify existing bugs and security issues in software projects.
It’s not surprising. AI-powered coding assistants sound awesome. But they are trained on existing code, and their suggestions reflect models of other programmers’ work – work that may be seriously flawed. Assistants’ assumptions create bugs that are often difficult to spot, especially when developers – who are adopting AI coding assistants in large numbers – rely on the assistants’ judgment.
In a less risky area beyond coding, Q Developer can help manage a company’s cloud infrastructure on AWS – or at least provide them with the information they need to manage it themselves.
Q Developer can respond to requests like “List all my Lambda functions” and “List my resources residing in other AWS Regions.” Currently in preview, the bot can also generate (but not execute) AWS CLI commands and answer questions related to AWS costs such as “What were the three most expensive services in the first quarter ? »
Image credits: Amazon
So, how much do these generative AI amenities cost?
Q Developer is available for free in the AWS Console, Slack, and IDEs such as Visual Studio Code, GitLab Duo, and JetBrains, but with limitations. The free version does not allow fine-tuning of custom libraries, packages, and APIs, and defaults to users in a data collection system. It also imposes monthly caps, including a maximum of 5 agent tasks (e.g. implementing a feature) per month and 25 requests on AWS account resources per month. (It baffles me that Amazon puts a cap on the questions one can ask about its own services, but here we are.)
The premium version of Q Developer, Q Developer Pro, costs $19 per month per user and adds higher usage limits, tools to manage users and policies, single sign-on and, perhaps most importantly , IP compensation.
Image credits: Amazon
In many cases, the models that underpin code generation services like Q Developer are trained on code that is copyrighted or licensed under a restrictive license. Vendors claim that fair use protects them in cases where designs were developed, knowingly or unknowingly, on copyrighted code – but not everyone agrees. GitHub and OpenAI are being sued in a class-action lawsuit that accuses them of violating copyright by allowing Copilot to regurgitate snippets of licensed code without providing credit.
Amazon says it will defend Q Developer Pro customers against claims alleging that the service violates a third party’s intellectual property rights, provided they let AWS control their defense and settle “as AWS deems appropriate.”
techcrunch
