Exploring the Synergy of Open Source Software and Generative AI Models (take one, and a half)

Exploring the Synergy of Open Source Software and Generative AI Models (take one, and a half)

May 12, 2023 from 11:00 am to 11:30 am

Speaker: Hans de Raad

A Factual Approach to Revolutionizing Software Development.

This conference talk explores the emerging advancements in software development, focusing on the intersection of open source software and generative AI models. Through an objective and factual lens, we will examine the opportunities and challenges these models present to software developers and development agencies.

Generational AI models are proving to be useful in various aspects of the software development lifecycle, thanks in part to their ability to encompass the vast body of knowledge available through open source code as training material. As a foundation for modern software, open source software facilitates the sharing of experience and knowledge, which actual usage adoption can be further amplified by developers integrating generative AI models into their daily development workflow.

The application of these models to generate the foundations for software can provide solutions to several issues, such as:

• License incompatibilities

• The “I am not a supplier” conundrum

• Dependency/supply chain management

Moreover, although slightly controversial, using generative AI models to generate code based on open source training code does uphold the spirit of open source collaboration, allowing developers to build upon each other’s work. In particular, these models can be applied to generic and repeatable software development tasks, generating code that alleviates the complexity of managing large numbers of dependencies.

Examples of generative AIs impact include the reduction of manual tasks in supply chain security scanning and quality assurance, which can be labor-intensive even when utilizing tools like Trivy and SonarQube. Generated code strategies make maintenance and updates easier by creating a new version with each project change. This will ensure that the dependency tree is stable and self-contained and will continue to be improved with new training data insights and security vulnerability mitigations upon each release.

Given the potential for generative AI models to reshape software development, this conference talk also addresses relevant laws and regulations, such as the European Union proposed Artificial Intelligence Act and the Cyber Resilience Act.

The integration of open source software and artificial intelligence models offers a promising avenue for transforming software development. By attending this conference talk, participants will learn more about this new technology and how it affects the industry.