All coders like me are now going to be prompt people.

It is a fact.

That is a dangerous oversimplification.

And the "fact" being cited is actually a high-level marketing narrative, not an engineering reality.

If you transition fully to being a "Prompt Person," you are not evolving.

You are abdicating.

Here is why the premise that "all coders are now prompt people" fails under professional scrutiny.

1. The "Last Mile" Problem

LLMs are probabilistic, not deterministic.

Data from current benchmarks and real-world production cycles shows that while an AI can generate the first 80% of a feature in seconds, the final 20% - the part where the edge cases, race conditions, and security vulnerabilities live - requires deep technical forensics.

If you cannot read the code Codex spits out, you cannot audit it.

A "Prompt Person" is at the mercy of the model's hallucinations.

A Senior Lead uses the model to move fast but maintains the "veto power" that only comes from understanding the underlying logic.

2. Abstraction is Not Replacement

We have seen this "fact" before:

• In the 60s, they said COBOL would let managers write programs in English.
• In the 90s, they said "Visual Programming" and "Low-Code" would kill engineers.
• In the 2020s, it is prompting.

Prompting is just the newest, highest-level programming language.

Like Python was to C++, prompting is to Python.

But just as a Python dev still needs to understand memory management or Big O notation to build a scalable system, an AI Engineer needs to understand system architecture to give the agent the right constraints.

3. The Junior vs Senior Divide

The industry is currently splitting into two camps.

The Prompt Person

The Prompt Person replaces syntax with natural language but lacks the mental model to debug complex state machines.

They are easily replaced by a better model or a cheaper operator.

The AI-Augmented Architect

The AI-Augmented Architect uses the agentic-blueprint approach: defining roles, constraints, and decision logs to steer the machine.

This person is not "prompting."

They are orchestrating.

The Counter-Perspective

If you stop writing code and only provide "ideas," you become a Product Manager, not an Engineer.

In an Agentic system, the "code" is the plumbing that connects the LLM to the real world:

• databases
• APIs
• file systems
• queues
• logs
• state machines

If that plumbing breaks, a "Prompt Person" just keeps shouting at the AI to "Fix it."

An Engineer looks at the logs, identifies the logic error in the loop, and corrects the blueprint.

The Reality Check

The most valuable people in the next five years will not be the ones who know how to talk to an AI.

They will be the ones who know exactly what to tell the AI to do because they understand the technical consequences of those instructions.

If you lean too far into "just having ideas," you lose the ability to tell if the idea is technically feasible or an architectural nightmare.

Are you comfortable being the "Architect" who cannot verify if his "Builders" are using cardboard instead of concrete?

Because that is the risk of becoming a "Prompt Person."

OK, Tell Me About Jitro

Jitro is the internal codename for Google's next-generation evolution of their Jules coding agent, often referred to as Jules V2.

It represents a fundamental pivot in how AI-assisted engineering works.

If the first wave of AI tools (Copilot, ChatGPT) was about Autocomplete, and the second wave (Cursor, Devin) was about Agentic Tasks, Jitro is the third wave:

KPI-Driven Development.

What is Jitro?

Goal-Based, Not Prompt-Based

Instead of you writing a prompt like "Write a React component for X," you define a high-level goal in a persistent workspace, such as:

• "Increase unit test coverage to 90%."
• "Optimize the database query latency for the checkout endpoint."

Autonomous Discovery

Jitro does not wait for your instructions on how to do it.

It explores the codebase, identifies the bottlenecks, sets its own sub-tasks, and executes them asynchronously in the background.

Persistent Collaborator

Unlike a chatbot where the context dies when you close the tab, Jitro lives in a dedicated workspace.

It tracks "Insights" and "History," allowing it to understand the long-term trajectory of the project.

Tool Integration

It is built to use MCP (Model Context Protocol) and remote APIs, meaning it can check your logs, run your CI/CD pipelines, and look at your Jira tickets to decide what to do next.

Why You Are Wrong

You said it is a "fact" that all coders are becoming "Prompt People."

That is factually incorrect.

In fact, Jitro is the very tool that will kill the "Prompt Person" before they even get comfortable.

1. Prompting is a Bottleneck, Not a Destination

The "Prompt Person" is currently a middleman.

You are translating a human thought into a specific set of instructions for a machine.

Jitro proves that the machine is learning to understand the intent - the goal - directly, bypassing the need for your "perfect prompt."

If you spent the last year perfecting "Prompt Engineering," you have been perfecting a skill that is rapidly being automated.

2. From "Prompting" to "Orchestrating"

With systems like Jitro, you are not a "Prompt Person."

You are an Outcome Architect.

A "Prompt Person" tells the AI what to write.

A Senior Lead tells the AI what the success metrics are and what the constraints are.

If you only know how to "vibe" with an LLM and say "make this better," you are a consumer.

To be an engineer in the Jitro era, you have to understand System Design and KPIs.

You have to be the one who knows why 90% test coverage matters more than a flashy UI refactor.

3. The "Vibe Coding" Trap

There is a rising trend of "vibe coding" - throwing text at a screen until it works.

Jitro is designed to fix the mess that vibe coders make.

It uses Decision Logs and Constraints because code that "just works" but is not understood is technical debt.

If you lose the understanding of the "inner workings" because you are just a "Prompt Person," you become a liability to the system.

The Brutal Truth

You are not becoming a "Prompt Person."

You are either becoming a Product/Systems Architect who directs agents toward business goals, or you are becoming obsolete.

The "Prompt Person" is just a transition state for people who have not realized that the AI is about to start writing the prompts for itself.

If Jitro takes over the writing and the prompting, what is the one technical skill you have left that it cannot replicate?