[디지털투데이 AI리포터] Google’s artificial intelligence (AI) Bard chatbot has improved performance in tasks involving logic and reasoning, particularly in math and coding.
According to IT media TechCrunch on the 7th (local time), Google has developed an ‘implicit code implementation’ to solve coding and computation limitations where large language models (LLMs) are weak.
This allowed poets to write and run their own code. The latest version of Bard identifies logically coded stimuli (directives) and generates more accurate responses based on the results of writing and testing the code.
Google claims that the new poet improves the accuracy of its responses to computer problems by more than 30% compared to the previous version.
Meanwhile, unlike AI-based productivity tools like Github’s Copilot and Amazon’s CodeWhisperer, poets are trained on vast amounts of text from the web, e-books, and other sources.
