High Code Quality Doesn't Matter to Machines
The Future Is Vibe Coding - Do We Like It Or Not
Hey, fellow Leader 🚀,
I am Artur, and welcome to my weekly newsletter. I am focusing on topics like IT Management, Innovation, and Leadership, with an Entrepreneurial mindset. My goal is to help you navigate the IT corporate landscape. Make better decisions, create awareness, and share real-world stories.
If you are leading an engineering team in 2026, you are likely handling a fascinating transformation.
In the pre-agentic era, the primary bottleneck in software engineering was code generation. Developers spent their energy writing code, translating logic into files, and trying to respect the team’s guidelines.
Today, AI agents have reduced the cost of creating code to almost a free experience.
But in doing so, we are watching a profound and historical transformation of the IT landscape. Especially on how we are building our new systems.
Technology No Longer Matters
Being an expert in a specific technology, writing code, migrating legacy systems, or writing simple unit tests are quickly becoming skills of the past.
Instead, the premium engineering skills of 2026 have shifted toward three core competencies:
System Architecture: The ability to design clean, highly decoupled boundaries between modules. Having a good starting architecture is now the primary prerequisite for AI leverage.
Context Engineering: The discipline of indexing and mapping the codebase so that AI agents have the exact, high-quality context they need.
Vibe Coding: Conversing naturally with our favorite agents is fast becoming the new standard for how we produce software.
Naturally, hardcore technical leaders complain about this shift toward vibe coding.
They argue that developers are losing agency over their own code, that repositories are being flooded with legacy debt, and that these systems are becoming incredibly difficult for a human to debug or maintain.
They are entirely right!
However, for the vast majority of business applications, vibe coding is proving to be more than enough.
High code quality was historically a human prerequisite to keep a codebase maintainable by other humans. It is not a machine prerequisite.
This is a monumental paradigm shift, and our mindset needs to adapt rapidly.
Juniors Don’t Need To Learn Code Anymore
This brings us to the biggest discussion point in modern engineering teams: how to properly train and onboard junior profiles.
Traditionally, junior engineers learned the craft by doing the “boring work”.
They wrote basic code, fixed simple bugs, and handled unit tests. Essentially, the tasks senior developers wanted to avoid. Under the guidance of mentors, they gradually leveled up their skills and system knowledge.
Today, AI does almost of that.
Consequently, companies have largely stopped hiring juniors, or they expect them to operate at a significantly higher level immediately by leveraging AI tools. Junior developers are no longer traditional developers. They are AI Developers now.
They are designing AI-oriented systems from their very first real-world projects.
This represents a massive opportunity (assuming they can manage to land a job in the first place).
If these juniors maintain a critical eye toward the systems they build, they will become the elite AI experts of tomorrow.
There is no longer a strict need to understand every line of code running behind a solution. Instead, the value lies in finding the perfect orchestration between AI tools, their integrations, and how they apply to a concrete business need.
In the future, developers will ultimately become technical business analysts.
It has been a wild ride since I published the first article on the SoW. Exchanging views with you and hearing feedback on how these articles have been useful is what drives my motivation to write. Leave a comment, subscribe to the SoW, and be part of the community.
If this article resonates with you, or you know someone who might find it useful, just share the link!
That’s it. If you find this post useful, please share it with your friends or colleagues who might be interested in this topic. If you would like to see a different angle, suggest it in the comments or send me a message.
Cheers,
Artur
We've actually seen something very similar on our team.
We had an idea for a feature that would normally go through the usual sequence: write the spec, review, plan the work, and then someone starts coding.
One engineer did almost all of that himself. He drafted the PRD with AI and then fed it into a coding agent, which came back with an implementation plan.
Writing the code ended up being the smallest part of the work. The attention went to reviewing the plan and deciding what was worth keeping .
What about embedded systems? IoT? Security-hardened systems?