Recently, Anthropic CEO Dario Amodei made headlines by saying that coding is going away first, followed eventually by software engineering. Predictably, social media exploded. Some people dismissed the statement as hype. Others declared the end of programming careers.
I believe his prediction is directionally correct.
Not because software is becoming less important—but because writing code is no longer the scarce skill it once was.
My Perspective After Three Decades in Software
I started building software in the mid-1990s.
Back then, writing code was the hard part.
You had to memorize APIs, understand obscure documentation, debug syntax errors, and spend hours reading manuals.
Today, AI can generate working code in seconds.
The bottleneck has shifted.
The difficult part is no longer writing software.
The difficult part is deciding what software should be built, how it should behave, how it fits into a business process, and how to ensure it produces reliable outcomes.
Coding Is Becoming a Commodity
History is full of examples where valuable skills became commodities.
Assembly language gave way to C.
C gave way to Java and Python.
Low-code platforms reduced the need for traditional programming.
Now AI is reducing the need to manually type code.
Each transition made developers more productive.
Each transition also eliminated jobs that focused only on the previous technology.
This wave is simply moving faster.
The Future Belongs to AI Systems Engineers
I don’t think software engineers disappear.
I think their role evolves.
Tomorrow’s engineers will spend more time on:
- Understanding business problems
- Product thinking
- Systems architecture
- Designing AI workflows
- Creating autonomous AI agents
- Integrating APIs and services
- Security and governance
- Human oversight and quality assurance
The keyboard becomes less important.
Judgment becomes more important.
The New Programming Language Is English
One of the biggest shifts we’re witnessing is that natural language is becoming a programming interface.
Instead of writing hundreds of lines of code, developers increasingly describe what they want.
AI produces the implementation.
This doesn’t eliminate engineering.
It changes what engineering means.
The engineer becomes the architect, reviewer, and orchestrator.
Business Knowledge Will Matter More Than Programming Knowledge
For many years, companies hired programmers because they could code.
Increasingly, they’ll hire people who understand an industry.
A healthcare expert using AI may outperform a brilliant programmer who doesn’t understand healthcare.
A logistics expert may build better AI systems than someone who only knows programming languages.
Domain expertise is becoming a competitive advantage.
What Should Students Learn?
If I were advising a student entering college today, I would still recommend learning programming.
But I would learn it differently.
Programming should be treated like mathematics.
You learn it to understand how computers think—not because you’ll spend your career typing code.
Alongside programming, I would invest heavily in:
- Critical thinking
- Problem solving
- Communication
- Product design
- Business processes
- AI tools and prompt engineering
- Agentic AI
- Automation platforms
- Systems thinking
These skills are much harder to automate.
What This Means for Software Companies
For software companies, this transformation is even bigger.
Clients won’t pay for code.
They’ll pay for outcomes.
The companies that thrive will be those that combine AI with deep business understanding to solve problems faster, cheaper, and better than before.
That requires fewer coders and more consultants, architects, AI orchestrators, and product thinkers.
Final Thoughts
I don’t see AI as the end of software engineering.
I see it as the biggest productivity revolution our industry has ever experienced.
Just as calculators didn’t eliminate mathematicians, AI won’t eliminate people who know how to solve meaningful problems.
But it will dramatically reduce the value of writing routine code.
The future belongs to those who can think clearly, understand businesses, collaborate with AI, and deliver results.
The age of coding is fading.
The age of AI-powered problem-solving has begun.
