Google’s Bard AI chatbot can now generate and debug code

Google’s conversation AI tool Bard can now help software developers with programming, including generating code, debugging and code explanation — a new set of skills that were added in response to user demand.

Coding has been one of the top requests Google has received from users, according to a Friday blog post by Google Research product lead Paige Bailey.

Google said Friday it is launching these software development capabilities in more than 20 programming languages including C++, Go, Java, JavaScript, Python and TypeScript. Users can export Python code to Google Colab. Bard can also help with writing functions for Google Sheets.

Collectively, this means that Bard, the generative AI experiment Google launched earlier this year, can review and help users debug their source code line by line. Google said developers can tell Bard “this code didn’t work, please fix it,” and it will help debug.

It can also translate code from one language to another and explain code snippets, a helpful feature for those new to programming.

Bard, which was created to compete with ChatGPT and other language models, didn’t quite stack up to its AI peers in TechCrunch’s own testing on chatbot performance. This latest set of skills could help Bard at least keep apace with ChatGPT and Claude — at least on paper.

How well Bard is able to create, translate and debug code is another question.

As Bailey cautioned in the blog post, Bard is still an early experiment and “may sometimes provide inaccurate, misleading or false information while presenting it confidently.”

For instance, Bard may give developers working code that is incomplete or doesn’t produce the expected output.

“Despite these challenges, we believe Bard’s new capabilities can help you by offering new ways to write code, create test cases, or update APIs. If Bard quotes at length from an existing open source project, it will cite the source,” Bailey wrote.