Google’s Bard AI Chatbot can Write and Generate Debug Code

Artificial Intelligence, Google Products, Technology

Listen to the Podcast:

You can open Table of Contents show

Bard, Google’s new AI chatbot, can now assist you in writing and debugging software code. Google announced the same via a blog post.

A few days ago, in an interview, the CEO of Google, Sundar Pichai, promised that their AI chatbot, Bard, would receive enhancements and improve over time. In light of the CEO’s statements, Google recently announced that Bard can now assist in writing and debugging code. In addition, the AI chatbot can explain code to users and assist with the creation of Google Sheets functions. Up until now, ChatGPT was known for assisting with code writing, but it appears that Google is eager to catch up. Bard is available to a small number of users in the United States and the United Kingdom and is still in testing.

Google Announces Bard’s New Powers

In the blog post announcing Bard’s new abilities, coding is described as ‘one of the top requests’ the company has received from its users.

The blog post describes Bard’s capabilities as follows: “As of now, Bard can assist with programming and software development duties, including code generation, debugging, and code explanation. More than 20 programming languages, including C++, Go, Java, Javascript, Python, and Typescript, will support these capabilities. And you can export Python code to Google Colab without having to copy and paste. Bard can also assist with Google Sheets writing functions.”

Adding that Bard can also assist in explaining code, the post states, “In addition to generating code, Bard can also assist in explaining code fragments. This is especially useful if you are learning to program for the first time, or if you require additional assistance to comprehend what a block of code may produce.”

On The Possibility of Hallucinating

We have heard a great deal about the AI chatbot’s tendency to ‘hallucinate’ at this juncture. AI is said to be hallucinating when it presents false information with assurance, leading the user to believe it to be true. The Google blog post about Bard includes a warning about the AI chatbot’s hallucination potential.

It states, “Bard is still an experiment in its infancy and may occasionally present inaccurate, misleading, or fraudulent information with assurance. Bard may provide you with functional code that does not produce the expected output, as well as code that is not optimal or incomplete. Before relying on Bard’s responses, always double-check them and thoroughly test and examine the code for errors, bugs, and vulnerabilities. Despite these obstacles, we believe that Bard’s new capabilities can assist you by providing you with new methods to write code, create test cases, and update APIs. If Bard cites extensively from an existing open-source project, the source will be cited.

What do Google Employees Think about Bard?

In February, when Bard had just been released, it was reported that employees were unhappy with the way the announcement was handled and were labeling it ‘rushed, botched, and un-Googley’. The company then solicited assistance from its employees to correct Bard’s errors, declaring incentives for those who agreed to assist. The CEO, Sundar Pichai, disclosed that more than 80,000 employees participated in the internal testing of Bard and even thanked them via email.

Bloomberg reports that more recently, Google employees who tested the AI chatbot Bard deemed it ineffective and pathologically dishonest. A worker also cautioned that Bard’s advice in certain circumstances could result in “death or injury.”

According to an internal document viewed by the publication, 18 current and former Google employees stated that Bard was disseminating low-quality information “in a race to keep up with the competition” while placing “less emphasis on its ethical commitments.”