Generative AI does present real societal risks, but not when it comes to Code Modernization

The editors at the mcode mansion flat take great interest in the risks posed by Generative AI, but they do not apply to what we do. We just liberate engineers to do their best work!

Generative AI raises potential risks that are mostly related to creating deepfake videos or disinformation, spreading propaganda, identity theft, unforeseen consequences of the decision-making process, etc. However, these concerns are largely irrelevant when it comes to using Generative AI for code modernization.

Code modernization tasks performed by Generative AI primarily involve code refactoring, translating legacy code into a modern language, or simplifying and optimizing existing code. Since this process deals with purely technical and systemic aspects, risks like spreading disinformation or creating deepfakes do not apply.

While there is a risk of generating incorrect or insecure code, these risks are not exclusive to generative AI. Human developers can make the same mistakes, sometimes even more frequently. However, this risk can be mitigated by proper testing and validation procedures.

Moreover, the AI model in this case is typically not exposed to external influences that could manipulate it to produce malicious output. It thus operates within a controlled environment, significantly reducing the potential for misuse.

Also, the human element remains because AI-generated code needs to be reviewed and approved by an experienced developer. This human audit minimizes the chance for errors or potential harmful code. So, the material risks associated with Generative AI in code modernization are extremely marginal compared to its use in other contexts.