Bard doesn’t have a worldwide release yet, but Google still seems to be continuously adding to the LLM’s capabilities.
Google has updated Bard to allow it to write code and debug programs. “Since we launched Bard, our experiment that lets you collaborate with generative AI, coding has been one of the top requests we’ve received from our users,” Google wrote in a blogpost.
Apart from writing code, Bard will be able to explain code snippets to users in English. Bard also debugs its own code when prompted. Google says Bard will also be able to help users optimize and improve their code. Bard will also cite the source if it takes a large chunk of code from an open source project.
Google, however, cautions that Bard’s code writing capabilities might not be perfect. “When it comes to coding, Bard may give you working code that doesn’t produce the expected output, or provide you with code that is not optimal or incomplete. Always double-check Bard’s responses and carefully test and review code for errors, bugs and vulnerabilities before relying on it,” Google says.
This new update brings Bard up to speed with GPT-4, which has already been doing much go this for a few months. Microsoft too has integrated GitHub Copilot within Github, and smaller companies like Replit are also coming up with similar software such as Ghostwriter that’ll help users write code.
Google, though, has some software creation tools of its own. Google Colab is one of the most popular platforms to write data science code, and Google could integrate Bard within it as a copilot. Google also operates the TensorFlow libraries, and could possibly use some of Bard’s capabilities there. It’s still early to predict how Bard will compare with GPT-4’s code writing capabilities, but at least Google now seems to be matching OpenAI in the features it offers in its LLMs.