A question about AI I've been carrying for a while

A developer questions whether programming languages should remain the primary interface between humans and computers in the age of large language models. They argue that current workflows still rely on translating human intent into code, and suggest that future systems might use representations designed for machines rather than humans. The developer emphasizes that understanding and verification will become even more critical as AI takes on more coding tasks.

I remember being at an advisory board of big tech in 2023, close to the starting point of code generation with LLMs. Like everyone else, I was amazed. People were copying and pasting generated code, experimenting with prompts, and imagining a future where AI could become every developer's pair programmer. I was excited too. But I remember having a completely different question. Not: -"Will AI write code?" Instead, I kept wondering: -Why are we still writing code at all? Not because I think programming is going away. Not because programmers won't be needed.And certainly not because programming languages are somehow "wrong." It was a much simpler question. Programming languages have always existed to bridge a gap between humans and machines. Humans think in goals, ideas, and intentions. Machines execute deterministic operations. Programming languages became the interface between those two worlds. For decades, we've improved that interface. Assembly became higher level languages. Higher level languages became frameworks. Frameworks became libraries and abstractions that let us think less about implementation and more about solving problems. Then large language models arrived. Suddenly, we could describe what we wanted in plain language. But instead of questioning the interface itself, we mostly asked AI to become incredibly good at translating our intentions into programming languages. Today, our workflow looks something like this: Human intent - LLM - Programming language - Compiler / Runtime -Machine execution And every time I look at this pipeline, I find myself asking the same question I had back then. -Are we optimizing the wrong layer? Maybe the question isn't "Can AI write code?" Maybe the question is: -Do programming languages still need to be the primary interface between humans and computers? This isn't an entirely new idea.Researchers have explored concepts like Intentional Programming, where software is represented by its intent rather than by a specific programming language. More recently, researchers have begun asking whether AI changes the role programming languages play in software development. But that's not exactly what I'm wondering. The question that keeps coming back to me is slightly different. Programming languages today are designed primarily for humans. They're readable. Maintainable. Reviewable.Collaborative. Those are wonderful qualities. But they're human qualities. Machines don't care whether a variable is named user or x. They don't care about formatting. They don't care about code style. They care about correctness, optimization, memory, scheduling, execution, and efficiency. So I can't help wondering: If we were inventing computing today, with powerful AI systems already available, would we still invent programming languages exactly as we know them? Or would we invent something entirely different? I'm not arguing that programmers should disappear. Actually, I think the opposite. As AI becomes more capable, I think understanding systems becomes even more important. Someone still needs to ask: Is this correct? Can we verify it? What assumptions is the model making? What happens when it fails? Who can debug the system if nobody understands what's happening underneath? Those aren't small questions but essential questions. Especially when software begins controlling physical systems like medical devices infrastructure, transportation or robots. AI can help us build incredible things. But trust doesn't come from generation. It comes from understanding. What if the next programming language isn't for us? This is the thought that has stayed with me the longest. Maybe the next major breakthrough isn't another language designed for humans. Maybe it's a computational representation designed primarily for machines. Not Python, not JavaScript ,not Rust. Something that naturally represents intent, constraints, goals, context, and relationships. Something that intelligent systems can reason about directly. Humans might never write that representation themselves. Just like most developers today rarely think about assembly. Maybe source code eventually becomes another abstraction layer still incredibly important, but no longer the primary interface. I don't know. But I think it's an interesting question to ask. Robotics makes this question even more interesting. Software lives in predictable environments. The physical world doesn't. Imagine telling a robot: -Pick up the glass without spilling the water, avoid the child running nearby, and place it gently on the table. That's not simply a sequence of instructions. It's intent, constraints, planning and continuous adaptation. Maybe future computational representations need to capture those ideas more naturally than today's source code does. And then I start thinking about biology. This is where my thoughts become even more speculative. AI is already helping researchers design proteins, analyze medical images, interpret genetic information, and develop brain-computer interfaces. Not because language models "understand biology" in the human sense. But because they can increasingly act as translators between human goals and incredibly complex systems. Could future computational representations make those interactions more natural? I honestly don't know. But it's a fascinating direction to think about. As excited as I am about AI, I also think we need to ask harder questions. Current large language models require enormous computational resources, Inference is expensive,training is even more expensive. What happens if every interaction with technology depends on massive models running in data centers? Is that sustainable? Can we build systems that are efficient enough for everyone? What happens when connectivity isn't available? What happens if energy becomes scarce? Or if access to these models is limited? We've spent decades building software that can run almost anywhere. I hope we don't lose that resilience. Maybe the future isn't replacing traditional software. Maybe it's finding the right balance between deterministic systems and probabilistic intelligence. When I think about that moment in 2023, I realize I wasn't imagining AI writing code. I was wondering whether code itself might eventually become another implementation detail,not something that disappears but something humans interact with less and less. Maybe programming languages remain one of humanity's greatest inventions, maybe they continue to evolve exactly as they always have or maybe, decades from now, we'll look at source code the same way most of us look at assembly today: an essential abstraction that few people work with directly. I don't know. And I'm not trying to predict the future. I'm simply sharing a question that's stayed with me for years. As AI models become more capable, robotics becomes more autonomous, and intelligent systems begin interacting with the physical world, I keep wondering whether we're optimizing the right layer of computing. Because maybe the most interesting question isn't: "Can AI write better code?" Maybe it's this: If we were inventing computers today, knowing what we know now, what language would we invent first and who would it be designed for? I'd love to hear what you think.