Stop Prompting AI: Why Anthropic's Boris Cherny Says the Future Is Loop Engineering | Mushood Hanif
Home / Blog / Stop Prompting AI. Start Building Loops. Why Boris Cherny Thinks the Future Is AI Orchestrating AI Insights
Stop Prompting AI. Start Building Loops. Why Boris Cherny Thinks the Future Is AI Orchestrating AI Boris Cherny, the head of Claude Code at Anthropic, recently revealed that he no longer prompts Claude directly. Instead, he builds loops that let AI agents decide, prompt themselves, evaluate their work, and continue autonomously. This isn't just a productivity hack—it's a glimpse into the next era of AI engineering.
Featured • loop engineering • June 28, 2026
Stop Prompting AI. Start Building Loops.
For the past three years, we've been obsessed with prompts.
Everyone wanted the perfect prompt.
Prompt libraries.
Prompt engineering.
Prompt marketplaces.
Prompt templates.
But according to Boris Cherny, the engineer leading Claude Code at Anthropic, that era is already ending.
"I don't prompt Claude anymore. I have loops running that prompt Claude and figure out what to do. My job is to write loops."
Read that again.
The person building one of the world's most capable AI coding assistants no longer spends his day chatting with AI.
He builds systems that let AI talk to itself.
That statement represents one of the biggest mindset shifts in AI engineering.
Prompt Engineering Was Never the End Goal
Prompt engineering solved a temporary problem.
Large language models were powerful, but they needed humans to tell them exactly what to do every single step of the way.
User
↓
Prompt
↓
LLM
↓
Answer
That workflow works well for:
brainstorming
writing
debugging
answering questions
But it falls apart when the task lasts for hours instead of minutes.
Real software projects don't end after one response.
Neither do research, design, testing, or code reviews.
The Rise of Loop Engineering Instead of asking AI one question after another, Boris builds loops .
A loop repeatedly asks four questions:
What's the goal?
↓
What's the next best action?
↓
Execute it.
↓
Did that move us closer to the goal?
↓
Repeat.
Notice something interesting.
The human isn't involved anymore.
The AI decides its own next prompt.
Prompts Become an Internal Implementation Detail This is the biggest conceptual shift.
The human no longer writes prompts.
Human
│
Goal
│
Orchestrator
│
──────────────
│ AI Planner │
│ AI Coder │
│ AI Reviewer│
│ AI Tester │
──────────────
Each AI generates prompts for the next AI.
Humans simply define objectives.
This Is How Claude Code Actually Feels If you've used Claude Code, Cursor, Codex, or Gemini CLI recently, you've probably noticed something.
You're writing fewer instructions.
Instead, you're giving goals.
The agent then decides to:
inspect the repository
understand the framework
create a plan
modify files
run tests
fix failures
review changes
repeat until complete
You didn't tell it every step.
AI Is Becoming Its Own Project Manager Think about a senior engineering manager.
open VS Code
create a file
write line 14
rename variable X
You give them an outcome.
They coordinate the work.
Loops bring that same abstraction to AI.
Instead of programming responses...
...you're programming workflows.
The Skills That Matter Next This changes what "AI engineering" actually means.
prompt writing
prompt optimization
clever wording
orchestration
state management
memory systems
evaluation loops
tool selection
planning
retries
verification
multi-agent coordination
The prompt hasn't disappeared.
It's just no longer the interface humans spend most of their time touching.
Why Loops Produce Better Results One-shot prompting assumes the first answer is good enough.
Real engineers know that's rarely true.
critique
refinement
retries
testing
validation
self-correction
The orchestrator can ask:
Did the tests pass?
Are there security issues?
Can performance improve?
Is the code idiomatic?
Should another agent review this?
The result is often dramatically better than a single prompt.
This Mirrors How Humans Work Good engineers don't produce perfect code on the first attempt.
Loops simply automate that cycle.
The AI isn't becoming smarter because of a magical prompt.
It's becoming more reliable because it's allowed to think in iterations.
The Future Isn't Prompt Engineering It's workflow engineering.
The winners won't necessarily be the people who write the cleverest prompts.
They'll be the people who build systems where AI agents can:
plan
execute
verify
recover
collaborate
improve continuously
The prompt is becoming infrastructure.
The workflow is becoming the product.
Final Thoughts Boris Cherny's statement isn't a rejection of prompt engineering.
They're just being written by software instead of humans.
If you've been learning prompt engineering, don't throw those skills away.
Instead, think one level higher.
"What's the best prompt?"
"What's the best system?"
That's the question defining the next generation of AI engineering.
