Large technology companies claim that artificial intelligence platforms like Microsoft Copilot can significantly improve productivity and efficiency for all levels of a business organization. However, it is important to determine whether these claims are exaggerated or have some validity.
For instance, can the free version of Microsoft Copilot, available to all Windows users, write a common SQL query upon command? This feature could be highly beneficial for a time-constrained and less experienced employee who is working towards a deadline. This article explores the capabilities of Microsoft Copilot in providing SQL assistance.
In a test conducted for this experiment, we began with a basic request to create a table in SQL called “Employee” with four columns: EMP_ID, Name, Salary, and Age. Microsoft Copilot successfully provided the code required for this task. Although some adjustments might be necessary based on specific requirements, the code provided was accurate and useful.
The explanation provided by Microsoft Copilot alongside the code is noteworthy, as it further clarifies the code’s purpose and functionality. Continuing the experimentation, we then asked Microsoft Copilot to generate the code for inserting a new record into the table. Once again, Copilot fulfilled this request accurately and helpfully.
Following this, we requested Copilot to generate code for listing employees from the table with a certain age or older. Microsoft Copilot suggested the appropriate code, which customers can customize according to their needs. This code proved to be accurate and useful, particularly for employees who are less familiar with SQL syntax.
Therefore, when properly prompted, Microsoft Copilot can generate basic SQL code for employees. However, users should always assess the generated answers for applicability and accuracy.
From this experiment, several conclusions can be drawn about Microsoft Copilot and its capabilities in SQL code generation:
1. Given the right conditions and questions, Copilot can provide useful and accurate SQL commands that employees can apply to their tasks.
2. The free version of Microsoft Copilot derives its “knowledge” from Bing searches and the internet. Consequently, asking Copilot about prevalent SQL commands with widespread availability on the internet is more likely to yield useful answers.
3. Relying solely on data from the internet for business decisions can be risky, as the internet often contains inaccurate or misleading information. Therefore, answers provided by Microsoft Copilot may be influenced by such inaccuracies in certain situations.
4. Even if the answers provided by Microsoft Copilot are useful, they should always be evaluated and filtered by employees for accuracy and applicability. While generative AI can be a powerful tool, it should not be wholly relied upon for making final decisions or acting autonomously.
5. The iterative nature of Microsoft Copilot sessions is one of its most powerful aspects. Previous answers can serve as a foundation for subsequent questions, allowing employees to engage in a conversation with the AI and achieve progressively more accurate and useful results.
6. The accuracy of the generated answer is influenced by the level of detail in the question. Including variable names enables Copilot to provide comprehensive answers instead of generic SQL command tips. Therefore, formulating detailed questions sets apart a simple search from a productive Copilot session.
7. Utilizing Microsoft Copilot requires a shift in mindset for employees who intend to use it as a workspace tool. It should not be approached as merely another search engine. Effectively, employees are requesting Copilot to read and interpret internet-available information to provide useful, accurate, and viable answers. The questions posed to Copilot will differ significantly from those asked of a basic search engine.
In conclusion, employees can trust Microsoft Copilot as a useful productivity tool for work tasks if they use it correctly. It is crucial to recognize that Copilot, like any other generative AI platform, is not simply a search engine. Questions must be well thought out, detailed, and specific. The more detailed the question, the more detailed the answer. Additionally, employees should understand that the initial question often serves as a foundation for an ongoing, valuable conversation with Copilot.
