Bard now uses a new technique called “implicit code execution” to recognize computing cues and run code in the background, the tech giant said in a blog post Wednesday.
Updated – Thu, 08 Jun 23 at 10:35am
San Francisco: Tech giant Google has introduced new improvements to its artificial intelligence (AI) chatbot Bard, including better logic and reasoning skills.
The tech giant said in a blog post on Wednesday that Bard now uses a new technique called “implicit code execution” to recognize computing cues and run code in the background.
As a result, it responds more correctly to string operations, encoding issues, and math operations.
“Our new method allows the Bard to generate and execute code to improve its reasoning and mathematical abilities. This approach was inspired by the well-studied dichotomy of human intelligence, in particular Daniel Kahneman’s book Thinking , Fast and Slow” — the separation of “System 1” and “System 2″ thinking,” it added.
System 1 thinking is fast, intuitive, and effortless. System 2 thinking, on the other hand, is slow, deliberate, and laborious.
With the latest update, the company combined the power of large language models (LLMs) (System 1) and traditional codes (System 2) to help improve the accuracy of Bard’s responses.
“We’ve seen this approach improve the accuracy of Bard’s answers to computation-based word and math questions in our internal challenge dataset by approximately 30 percent,” the tech giant said.
Additionally, the company announced that Bard now offers a new export option for Google Sheets.
With the new feature, users can export tables that Bard creates in their responses directly into Sheets.
