Learn about CodeWhisperer, Amazon’s AI-powered assistive coding tool. As of today it is a kind of kaput.
CodeWhisperer is now a Q Developer, part of Amazon’s Q family of business-focused generative AI chatbots, expanding with the newly announced Q Business. Available through AWS, Q Developers are here to help. As CodeWhisperer did, it includes some of the tasks developers perform in the course of their daily work, such as debugging and upgrading apps, troubleshooting issues, and performing security scans.
In an interview with TechCrunch, Doug Seven, GM and Director of AI Developer Experience at AWS, implied that CodeWhisperer was a bit of a branding failure. Third-party metrics reflect as much.; Despite having a free tier, CodeWhisperer has struggled to keep up with the momentum of its main rival, GitHub Copilot, which has more than 1.8 million paying individual users and tens of thousands of enterprise customers. (Poor initial impressions certainly didn’t help.)
“CodeWhisperer is where we started (with code generation).But we wanted to have a brand and name that fit a wider range of use cases,” Seven said. “You may think Q Developer is an evolution of CodeWhisperer into something much broader.”
To achieve this, Q Developer not only allows you to generate code including SQL, a programming language commonly used to create and manage databases, but also helps you test that code and transform and implement new code inspired by developer queries. can give.
Like Copilot, customers can fine-tune Q Developer on their internal codebase to improve the relevance of the tool’s programming recommendations. (The now-deprecated CodeWhisperer also offered this option.) And thanks to features called agents, Q Developer can autonomously perform tasks such as implementing features, documenting code, and refactoring (i.e. restructuring).
You can make a request to Q Developers, such as “Please create an ‘Add to Favorites’ button in my app”, and Q Developers will analyze your app’s code, generate new code if necessary, create a step-by-step plan, and complete testing. Write code before implementing proposed changes. Developers can review and iterate over the plan before Q implements it, linking steps together and applying updates to necessary files, code blocks, and test suites as a whole.
“What happens behind the scenes is that Q Developer actually spins up the development environment to work on the code,” Seven said. “So for feature development, Q Developer takes an entire code repository, creates a branch of that repository, analyzes the repository, performs the requested operation, and returns those code changes to the developer.”
Image Credits: Amazon
Additionally, the agent can automate and manage the code upgrade process, as Java conversions (specifically from Java 8 and 11 built using Apache Maven to Java version 17) are live in real time, and .NET conversions are coming soon, Amazon said. says: Seven added, “Q Developer analyzes the code to find what needs upgrading and makes any changes before returning it to the developers.”
To me, Agents sounds very similar to GitHub’s Copilot Workspace, which similarly creates and implements plans for bug fixes and new features in the software. And like Workspace, I’m not entirely convinced that this more autonomous approach can solve the problems associated with AI-based coding assistants.
GitClear’s analysis of more than 150 million lines of code committed to project repositories over the past few years found that Copilot causes more bad code to be pushed into the code base. Elsewhere, security researchers have warned that Copilot and similar tools could amplify existing bugs and security issues in software projects.
This is not surprising. The AI-based coding assistant looks impressive. But they are trained on existing code, and their suggestions reflect patterns of other programmers’ work, work that may have serious flaws. The assistant’s guesswork often creates bugs that are difficult to find, especially when developers who mass adopt AI coding assistants rely on the assistant’s judgment.
In less risky areas outside of coding, a Q Developer can help you manage your company’s cloud infrastructure on AWS, or at least give you the information you need to manage it yourself.
Q Developer can fulfill requests such as “list all Lambda functions” and “list my resources in different AWS Regions”. Currently in preview, this bot can generate (but not execute) AWS Command Line Interface commands and answer AWS cost questions, such as “What were the top three most expensive services in Q1?”
Image Credits: Amazon
So what is the cost of these generative AI conveniences?
Q Developer is free to use in the AWS Console, Slack, and IDEs such as Visual Studio Code, GitLab Duo, and JetBrains, but there are limitations. The free version does not allow fine-tuning of custom libraries, packages, and APIs, and defaults to choosing the user as the data collection scheme. We also impose monthly limits, including up to 5 agent operations (e.g. feature implementations) per month and 25 queries to AWS account resources per month. (It’s baffling to me that Amazon places limits on the questions you can ask about its own services, but that’s it.)
Q Developer Pro, the premium version of Q Developer, costs $19 per user per month and adds higher usage limits, user and policy management tools, single sign-on, and perhaps most importantly, IP indemnification.
Image Credits: Amazon
In many cases, the models behind code generation services like Q Developer are trained on code that is copyrighted or under a restrictive license. Vendors argue that fair use protects them if their models are knowingly or unknowingly developed based on copyrighted code, but not everyone agrees. GitHub and OpenAI are being sued in a class action lawsuit accusing Copilot of violating copyright by allowing licensed code snippets to be regurgitated without providing credit.
Amazon said it will defend Q Developer Pro customers against claims alleging that the service infringes the IP rights of third parties, as long as AWS controls its defenses and allows them to be resolved “as AWS deems appropriate.”
