Using generative artificial intelligence (AI) solutions to produce computer code helps streamline the software development process and makes it easier for developers of all skill levels to write code. The user enters a text prompt describing what the code should do, and the generative AI code development tool automatically creates the code. It can also modernize legacy code and translate code from one programming language to another.
By infusing artificial intelligence into the developer toolkit, these solutions can produce high-quality code recommendations based on the user’s input. Auto-generated code suggestions can increase developers’ productivity and optimize their workflow by providing straightforward answers, handling routine coding tasks, reducing the need to context switch and conserving mental energy. It can also help identify coding errors and potential security vulnerabilities.
Generative AI for coding is possible because of recent breakthroughs in large language model (LLM) technologies and natural language processing (NLP). It uses deep learning algorithms and large neural networks trained on vast datasets of diverse existing source code. Training code generally comes from publicly available code produced by open-source projects.
Programmers enter plain text prompts describing what they want the code to do. Generative AI tools suggest code snippets or full functions, streamlining the coding process by handling repetitive tasks and reducing manual coding. Generative AI can also translate code from one language to another, streamlining code conversion or modernization projects, such as updating legacy applications by transforming COBOL to Java.
Even as code produced by generative AI and LLM technologies becomes more accurate, it can still contain flaws and should be reviewed, edited and refined by people. Some generative AI for code tools automatically create unit tests to help with this.
Using AI code generation software is generally straightforward and available for many programming languages and frameworks, and it’s accessible to both developers and non-developers.
There are three main benefits of using AI code-generation software tools:
It saves time by enabling developers to generate code faster, reducing the work of manually writing lines of code and freeing developers to focus on higher-value work. Generative AI can quickly and efficiently test and debug computer code. Using generative AI for code also makes code development accessible to non-developers.
Generative AI, low-code and no-code all provide ways to generate code quickly. However, low-code and no-code tools depend on prebuilt templates and libraries of components. The tools enable people without coding skills to use visual interfaces and intuitive controls like drag-and-drop to create and modify applications quickly and efficiently while the actual code remains hidden in the background. Generative AI for code software, on the other hand, doesn’t use templates and libraries of components. The software reads a developer’s plain language prompts and suggests code snippets from scratch that will produce the desired results.
While low-code and no-code tools generally target non-developers and business users, both professional developers and other users can use AI code-generation software.
The Team4 Code Whisperer works as an extension to Visual Studio Code.
Download the Visual Studio Code extension | Updated 04 July 2024, 941 MB
The Visual Studio Code IDE extension support the following operating systems and hardware architectures:
The model used is the IBM Granite 20B enterprise Java code instruct model. This model was trained in two phases. In the first phase, the model was trained on 3 to 4 trillion tokens sourced from 116 programming languages, ensuring a comprehensive understanding of programming languages and syntax. In the second phase, the model was additionally trained on 500 billion tokens with a carefully designed mixture of high-quality data from code and natural language domains to improve the model's ability to reason. For more information about the IBM Granite 20B enterprise Java code instruct model, see Granite Code Models: A Family of Open Foundation Models for Code Intelligence.
Training of the IBM Granite 20B enterprise Java code instruct model included datasets for Jakarta EE, MicroProfile, and other enterprise Java repositories such as Open Liberty GitHub repositories.
Generative AI provides various explanations, including the following ones: