Are We Asking the Wrong Question About AI Replacing Programmers?

Will AI can replace programmers ? The Real Question is will humans ever stop seeking more precise expression?

I believe we humans always strive to express ideas in a more precise form. This requires a deeper contextual understanding of the system. As computers advance, our ability to describe and manipulate systems grows deeper. In my opinion,

Role of Programming Languages in Methodological Expression

In my view, the book SICP: Structure and Interpretation of Computer Programs has proven that AI cannot replace programmers, even 40 years ago. This is, of course, a personal viewpoint, open to be proven wrong.

"First, we want to establish the idea that a computer language is not just a way of getting a computer to perform operations but rather that it is a novel formal medium for expressing ideas about methodology." — SICP

The Evolution of Computing and the Endurance of Technical Depth
The history of computing supports this notion. Abstraction layers have not rendered deeper technical understanding obsolete; they have, instead, co-evolved with fundamental advancements in computing.

Coexistence of High-Level and Low-Level Programming

Virtual machine (VM)-based programming languages like Java, C#, Python, and JavaScript have not replaced the need for system-level programming—instead, they coexist with low-level languages like C, C++, and Rust, which remain critical not only for performance but also for operating system kernels, system security, and hardware interaction.

Co-evolution of Programming Languages and Hardware Architectures.

As C and C++ pushed for more control and efficiency, ISAs evolved to unlock new possibilities, driving performance and scalability to levels once thought impossible. And as architectures like RISC streamlined instructions and CISC embraced complexity, compilers evolved to squeeze out every ounce of efficiency.

This dynamic is a feedback loop of innovation where software and hardware continuously push each other forward. In the end, it's this synergy that fuels the exponential growth of modern computing, reminding us that real breakthroughs happen when disciplines collide and evolve together.

The need to express precise low-level computation instructions ensures that ISAs—or something better—will keep evolving. As hardware demands shift with advancements in AI, quantum computing, and specialized accelerators, rigid instruction sets won’t be enough. The focus will move toward more flexible, efficient, and domain-specific ways to control hardware.

RISC-V is breaking away from traditional architectures like CISC due to the ongoing tension between complexity, efficiency, and the demand for flexibility in modern computing. CISC architectures (like x86) were designed to handle complex instructions within a single cycle, but this led to increased hardware complexity and less adaptability in a rapidly evolving tech landscape.

The Continued Importance of Foundational Knowledge in Circuit Design.

While VHDL and Verilog have streamlined digital circuit design, a deep understanding of digital logic remains essential. Engineers with strong fundamentals can debug efficiently, optimise for performance, and minimise resource usage in FPGA and ASIC designs. Knowledge of Boolean algebra, Karnaugh maps, and FSMs ensures better synthesis and hardware efficiency. Foundational skills also help in adapting to new technologies, bridging theory with real-world implementation. Despite advancements in HDL, digital logic remains the backbone of efficient and reliable circuit design.

Semiconductor Fundamentals Still Matter in Modern Digital Circuit Design.

While digital logic abstractions like logic gates and finite state machines have simplified circuit design, they fundamentally rely on the behavior of underlying semiconductor devices. A deep understanding of transistor-level physics remains essential for optimizing performance, power efficiency, and reliability. Technologies like CMOS, which dominate modern circuits, require knowledge of charge dynamics, leakage currents, and material properties to manage power consumption and thermal performance effectively. Advanced architectures such as FinFETs and GAAFETs, along with considerations of signal integrity and noise immunity, stem directly from semiconductor innovations rather than logical abstractions. Designers must also account for process variations and fabrication constraints that influence circuit behavior. As devices scale to the nanoscale, quantum effects further impact the behavior of logic gates, making semiconductor knowledge critical. Ultimately, mastering these fundamentals is vital for pushing the boundaries of digital design and achieving greater efficiency and performance.

Progressing Beyond Classical Electronics

While classical electronics and semiconductor physics form the backbone of modern technology, they haven’t limited innovation—they’ve propelled it. Engineers and scientists continue to push beyond traditional boundaries, diving into quantum mechanics to explore new computational possibilities and improve material properties. Advances in fabrication processes, such as EUV lithography and 3D chip stacking, have enabled more efficient, powerful, and compact devices. Meanwhile, quantum principles are increasingly influencing chip design, paving the way for quantum computing, spintronics, and neuromorphic architectures. By integrating classical principles with emerging technologies, the evolution of electronics continues to accelerate, redefining what’s possible in modern computing.

"The computer revolution is a revolution in the way we think and in the way we express what we think. The essence of this change is the emergence of what might best be called procedural epistemology—the study of the structure of knowledge from an imperative point of view, as opposed to the more declarative point of view taken by classical mathematical subjects." — SICP

Programming Languages as Tools for Procedural Epistemology
A programming language is a formal system for expressing procedural epistemology, providing a precise and structured way to define, manipulate, and execute knowledge. Unlike natural languages, it is rigorous and unambiguous, enabling clear representation of computational processes.