Google Bard was designed as a free AI service launched for people of the internet world, which could generate factual information and provide relevant answers, which can be helpful for research purposes. But it was observed in the initial phases that Some of its responses may lack originality or may not directly address the questions asked. Upon its initial launch, Bard faced substantial criticism, particularly regarding its reasoning capabilities, prompting individuals to draw comparisons with OpenAI’s ChatGPT.
According to a recent blog post by Google, its AI-powered chatbot, Bard, is gradually making progress in tasks related to logic and reasoning. The post highlights the implementation of “implicit code execution” as a technique that has led to specific improvements in Bard’s performance in math and coding domains.
The blog further explained that Bard is a long-text generating software that’s why, Large language models (LLMs) like Bard function as prediction engines. Their response generation is based on anticipating the most probable words to follow a given prompt. While they excel at tasks like email and essay writing, they may face challenges when it comes to software development due to their tendency to make errors.
According to Google’s internal benchmarking, the latest version of Bard has demonstrated a significant improvement of 30% in its responses to “computation-based” word and math problems compared to the previous release. However, it remains to be seen whether these claims will withstand external testing and scrutiny.
Bard product lead Jack Krawczyk and VP of engineering Amarnag Subramanya wrote in the blog post,
Even with these improvements, Bard won’t always get it right — for example, Bard might not generate code to help the prompt response, the code it generates might be wrong or Bard may not include the executed code in its response.
Google’s claim regarding Bard’s improvement in math and programming represents a remarkable milestone in the field of AI. As Bard continues to evolve and refine its capabilities, it opens up exciting possibilities for the integration of AI technology in various domains.
