Google’s AI System Improving its Programming Skills
In the continuous pursuit of having the most sophisticated AI, it appears that a new announcement is made on a daily basis. Google has recently revealed its latest update to Bard, which includes enhanced advanced reasoning and mathematical capabilities.
Bard no longer relies exclusively on LLMs who are best at forecasting versus solving complex problems. Instead, it should now recognize when further processing could help, and then generate code-behind to improve correctness. According to Google, this update improved the accuracy of calculus-based word and math problems in their internal challenge datasets by 30 percent.
Google claims that Bard should be able to provide information such as the main factors behind millions of numbers, how fast your savings are growing, and what words like “lollipop” look like backwards (as it will no doubt convince AI naysayers that it’s the ultimate tool). The company itself provides a standard disclaimer explaining: “Bard doesn’t always get it right – for example, Bard may not generate code to help with a quick response, the code it generates may be incorrect, or Bard may not include executed code in its response.”
In addition to the integration of code generation, Google has announced a function that allows the data produced by Bard to be edited and processed in Sheets. It’s a small update, but people who hate spreadsheets will no doubt appreciate it. It’s a small update, but it takes some of the stress out of building and maintaining spreadsheets.
