From Artificial Intelligence To Artificial Wisdom

In the article "From Artificial Intelligence To Artificial Wisdom," Itamar Friedman highlights the evolving landscape of AI in software development, asserting that while AI has revolutionized code generation speed, the new critical challenge lies in validating, governing, and trusting the code produced by AI. Friedman introduces the concept of "artificial wisdom" as the necessary progression to ensure the long-term viability and sustainability of the speed gained from AI. Artificial wisdom is characterized by two key attributes: being stateful and specific. A stateful system maintains an ongoing model of an organization's standards, incorporating insights from developer interactions, historical code, and review feedback. This allows governance to build and evolve rather than being reset with each new task. Moreover, artificial wisdom is specific, meaning it enforces an organization's unique standards and risk tolerances, moving beyond generic industry best practices. This tailored approach enables organizations to confidently manage the quality of the expanding volume of AI-generated code within their specific operational context. The article concludes by urging engineering leaders to adopt a continuous and context-aware verification process, establishing governance systems that consistently apply accumulated organizational judgment to all AI-generated code.