Google’s efforts in the generative AI field withBardhave been minimized by therapid rise of OpenAI’s ChatGPT, which is also featured on Microsoft’s Bing Mobile and Skype apps. In fact,a December 2022 reportsuggested that Google was issuing a “code red” internally when ChatGPT was made available to the masses. ChatGPT has supported coding for a while now, but Bard was late to the party. Thatchanged last month, and the company is now expanding AI coding capabilities to Google Colab, its cloud-based Python development environment.

Among the new AI features coming to Colab is a dedicated chatbot to help you through the coding process, plus the ability to generate code using natural language. Inits announcement post, Google says these features come to Colab through Codey, a generative AI model focusing on coding. It is based on PaLM 2 and was detailed at the recently concludedI/O 2023.

Google Colab Integrated Chatbot-anim

In its initial phase of availability. Codey models within Colab will appear for users based in the US. Google says the first set of features will “focus on code generation,” adding that the new integration can “dramatically increase programming speed, quality, and comprehension.” This new feature rollout also eliminates the need to type the same type of code repeatedly. When available, “eligible” Colab users will notice aGeneratebutton and a box next to it for text prompts, as illustrated above.

Then there’s a new chatbot making its way to Colab, though its rollout is listed as “soon,” with no definitive timeline available. Google says it can respond to basic queries like “How do I import data from Google Sheets?” or “How do I filter a Pandas DataFrame?” to further enhance the AI experience for beginners and professionals alike.

Google Colab Autocomplete for paid subscribers

This new solution saves the trouble of generating code on Google’s chatbot and exporting it to Colab. It is important to remember that any code generated via prompts on a chatbot could be used by Google to improve its models. So short of satisfying your curiosity, this method is not recommended for code you’d rather not have publicized.