Google Bard, the famous ChatGPT rival has been improving almost daily. With already an upper hand to provide real-time information, Bard is now getting a new export feature added to the platform along with abilities to perform better at mathematics and coding. Here is everything you need to know about the latest development of the platform.
How is Google Bard Improving?
In recent months, Google has been making significant improvements to Bard's ability to perform mathematical and coding tasks. These improvements are being made through a combination of techniques, including:
Implicit code execution: This technique allows Bard to generate code that can be executed by a computer. This means that Bard can now solve more complex mathematical problems and write more sophisticated code.
The Google blog says “A new technique called “implicit code execution” helps Bard detect computational prompts and run code in the background. As a result, it can respond more accurately to mathematical tasks, coding questions, and string manipulation prompts.”
Improved and better reasoning: Being a Large Language Model (LLM) Bard can predict texts which is helpful for creative work however this means it lacks the ability for logical reasoning. That is why the new updates on the platform will allow it to understand the relations between different concepts.
The official statement reads “Large language models (LLMs) are like prediction engines — when given a prompt, they generate a response by predicting what words are likely to come next. As a result, they’ve been extremely capable on language and creative tasks, but weaker in areas like reasoning and math.”
“In order to help solve more complex problems with advanced reasoning and logic capabilities, relying solely on LLM output isn’t enough. Our new method allows Bard to generate and execute code to boost its reasoning and math abilities”, it added.
To test its new abilities, a prompt stating “Create a Microsoft Excel Sheet on the top 10 world's most populated countries ” was presented and this was Bard’s response.
“Sure, here is an Excel sheet for the top 10 world's most populated countries:”
| Country | Population | Yearly Change | Net Change | Land Area (Km²) | World Share |
| China | 1,444,216,107 | 0.04% | 62,206 | 9,597,000 | 18.47% |
| India | 1,412,069,323 | 0.99% | 13,586,631 | 3,287,263 | 17.70% |
| United States | 332,459,491 | 0.59% | 1,937,734 | 9,833,520 | 4.25% |
| Indonesia | 274,221,326 | 1.07% | 2,898,047 | 1,919,443 | 3.51% |
| Pakistan | 222,527,818 | 2.00% | 4,327,022 | 880,940 | 2.83% |
| Brazil | 213,317,693 | 0.72% | 1,509,890 | 8,511,969 | 2.73% |
| Nigeria | 207,301,429 | 2.58% | 5,175,990 | 910,770 | 2.64% |
| Bangladesh | 165,848,103 | 1.01% | 1,643,222 | 130,170 | 2.11% |
| Mexico | 129,530,753 | 1.06% | 1,357,224 | 1,943,950 | 1.65% |
Bard generated data for only 9 countries, this shows that the platform is currently not accurate. However, Google warns users on this “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.”
In conclusion, the possibilities for Google Bard are truly endless. It is an exciting new technology that has the potential to change the world in many ways. However, the AI model still needs a lot of development to give accurate results.
Comments
All Comments (0)
Join the conversation