Give a gift to CodeWhisperer, Amazon’s AI-powered assistive coding tool. To this day it is broken – somehow.
CodeWhisperer is now Q Developer, part of Amazon’s Q family of business-focused generative AI chatbots, which also extends to the newly announced Q Business. Available via AWS, Q Developer helps with some of the tasks that developers do in the course of their daily work, such as debugging and updating apps, troubleshooting, and running security scans – similar to what CodeWhisperer did.
In an interview with TechCrunch, Doug Seven, GM and Director of AI Developer Experiences at AWS, suggested that CodeWhisperer was something of a branding mistake. Third-party metrics reflect this; Even with a free tier, CodeWhisperer struggled to keep up with the momentum of its main competitor GitHub Copilot, which has over 1.8 million individual paying users and tens of thousands of corporate customers. (Bad early impressions certainly didn’t help.)
“We started with CodeWhisperer [with code generation]b“But we really wanted to have a brand – and a name – that fit a broader range of use cases,” Seven said. “You can imagine Q Developer as an evolution of CodeWhisperer into something much more comprehensive.”
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 assist in the transformation and implementation of new code derived from developer requests.
Similar to Copilot, customers can use Q Developer to tweak their internal codebases to improve the relevance of the tool’s programming recommendations. (The now-deprecated CodeWhisperer also offered this option.) And thanks to a feature called Agents, Q Developer can independently do things like implement features and document and refactor (i.e., restructure) code.
Submit a request to Q Developer 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 run tests on the code before making the suggested changes. Developers can review and iterate the plan before Q implements it by connecting steps and applying updates to the required files, code blocks, and test suites.
“What happens behind the scenes is that Q Developer actually 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, analyzes the repository, does the work assigned to it, and returns those code changes to the developer.”
Photo credit: Amazon
Amazon says agents can also automate and manage code update processes, with Java conversions live today (particularly Java 8 and 11 built with Apache Maven on Java version 17) and .NET conversions coming soon. “Q Developer analyzes the code – looking for anything that needs to be updated – and makes all of those changes before handing it back to the developer to review and commit,” Seven added.
To me, “Agents” sounds a lot like GitHub’s Copilot Workspace, which also generates and implements plans for bug fixes and new features in software. And – like Workspace – I’m not entirely convinced that this more autonomous approach can solve the problems associated with AI-powered coding assistants.
An analysis of over 150 million lines of code committed to project repos from GitClear over the past few years found that Copilot resulted in more broken code being pushed into codebases. Elsewhere, security researchers have warned that Copilot and similar tools can exacerbate existing bugs and security issues in software projects.
That’s not surprising. AI-powered coding assistants seem impressive. But they are trained on existing code, and their suggestions reflect patterns in other programmers’ work – work that can have serious flaws. The assistants’ guesses lead to errors that are often difficult to detect, especially when developers – who use AI coding assistants in large numbers – defer to the assistants’ judgment.
In a less risky area that goes beyond coding, Q Developer can help manage a company’s cloud infrastructure on AWS – or at least get it the information it needs to do the management itself.
Q Developer can fulfill requests such as “List all my Lambda functions” and “List my resources located in other AWS Regions”. Currently in preview, the bot can also generate (but not execute) AWS Command Line Interface commands and answer questions about AWS costs, such as “What were the three highest-cost services in the first quarter?”
Photo credit: Amazon
How much do these generative AI amenities cost?
Q Developer is available for free in the AWS console, Slack, and IDEs like Visual Studio Code, GitLab Duo, and JetBrains – but with limitations. The free version does not allow for fine-tuning of custom libraries, packages, and APIs and selects users for a data collection scheme by default. Monthly limits are also set, including a maximum of 5 agent tasks (e.g., feature implementation) per month and 25 AWS account resource queries per month. (It’s baffling to me that Amazon would limit the questions you 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 offers higher usage limits, user and policy management tools, single sign-on, and – perhaps most importantly – IP compensation.
Photo credit: Amazon
In many cases, the models underlying code generation services such as Q Developer are trained on code that is copyrighted or subject to a restrictive license. Vendors claim that fair use protects them in the event that the models were knowingly or unknowingly developed on copyrighted code – but not everyone agrees. GitHub and OpenAI are being sued in a class action lawsuit accusing them of violating copyright law by allowing Copilot to reproduce licensed code snippets without attribution.
Amazon says it will defend Q Developer Pro customers against claims alleging that the service infringes the IP rights of third parties as long as they let AWS control their defense and reach a settlement “as AWS deems appropriate.” .