AI Software Design: A New Era for Innovation
Exciting times are ahead as I pivot from a fulfilling job to dive deep into AI software design. My research aims to tackle the challenges faced by AI in creating high-quality, large-scale applications.
Thesis:
- AI excels at writing small programs but struggles with bigger, more complex tasks.
- The software industry often treats AI merely as an oracle, overlooking its potential as a foundational design concept.
Research Areas:
-
AI-First Design Patterns:
- Discovering simple patterns, like “determinism invalidation,” to ensure that AI-generated code remains relevant even as environments change.
-
High-Quality Software Creation:
- Bridging the gap between AI’s capability and real-world software challenges, focusing on specificity and flexibility.
-
Modeling AI from First Principles:
- Rethinking how AI tools are constructed for better integration and performance.
What comes after monoliths? The future involves custom-built tools driven by AI, tailored to specific needs.
👉 Let’s connect! Share your thoughts on AI’s future in software design below.
