Google introduces the ability to generate, debug and explain code in the ‘chatbot’ Bard

21 Apr. (Portaltic/EP) –

Google has updated his chatbot bardpowered by Artificial Intelligence (AI), introducing the ability to help users with programming tasks and software development, with functions for generate, debug, and explain code.

The ‘chatbot’ developed by Google, which is based on the Language Model for Dialogue Applications conversation technology (LaMDA), was introduced in February with the intention of enhancing the search experience and offering “current, high-quality answers” as well as collaborating with generative AI.

Now the tech giant has announced a new bard upgrade with which it offers help users to carry out tasks programming and software developmentwith the aim that everyone can program, “with whatever level of experience they have”.

In this sense, as explained by Google in your blogusers will be able ask Bard to generate code, it debug or develop explanations about how a particular code works. Likewise, it has also indicated that this capacity will be launched with more than 20 programming languagesincluding some of the more popular ones like C++, java, javascript, python and Typescript.

Another feature that Google highlights is that Bard allows export the code that has been written with Python in a simple way, since it allows you to use Google Colab, to export it “without the need to copy and paste”. In addition, this function of Bard has also been introduced in writing functions for Sheets of Google.

Regarding the ability to explain codeGoogle has indicated that it is a function focused on those users who are learning programming for the first time, as well as to help them understand “what a block of code might generate”.

On the other hand, Bard also performs a code debugging process for identify if this has any errors. In this case, Bard gives an error message and you may be prompted to fix it. It can also help the user to make a existing code faster and more efficient.

Despite all this, Google has stressed that Bard is “an early experiment” and that, therefore, sometimes may provide “inaccurate, misleading or false” information. In this sense, it warns that the ‘chatbot’ can make mistakes such as providing a functional code that does not produce the correct result or that it is incomplete and, therefore, recommends checking the code “carefully” before “trusting” it.