Recently, I came across an interesting idea while reading a book about algorithms and systems thinking.
It described Elon Musk’s approach to improving productivity and processes.
Surprisingly, the framework is very simple:
- Question whether the task is necessary.
- Simplify the process.
- Automate only as the final step.
At first glance, this sounds obvious.
But when you look at how most companies operate, many do exactly the opposite.
Most Companies Start With Automation
A process becomes slow.
Immediately people suggest:
“Let’s automate it.”
“Let’s add AI.”
“Let’s build a new system.”
Today the solution is often:
“Let’s use AI agents to handle it.”
But Musk would ask a different question first:
“Why does this process exist in the first place?”
That question is often far more important than the choice of software.
The Best Process Is the One That Doesn’t Exist
Imagine a company requires approval from:
- An engineer
- A manager
- A director
- An executive
before a decision can be made.
Most organizations will try to improve the approval workflow.
They build a digital approval system.
They add mobile notifications.
They integrate AI-generated summaries.
They automate reminders.
Musk’s approach is different.
He first asks:
“Why do we need all four approvals?”
“Can we remove one?”
“Can we remove three?”
“Do we even need this report?”
If a process can be eliminated, that is usually a much bigger productivity gain than automating it.
After all, no process is faster than a process that no longer exists.
Simplification Before Automation
The second principle is simplification.
Suppose a report requires:
- 20 inputs
- 10 calculation steps
- 5 approval stages
Most companies simply digitize all 35 steps.
Musk would first examine whether those 35 steps should exist at all.
Perhaps 20 inputs can become 8.
Perhaps 10 calculations can become 3.
Perhaps 5 approvals can become 1.
Only after that simplification would he consider automation.
The Trap of Automation: Japan’s Stamp Robot
This reminds me of one of the most famous examples of process automation.
Japan’s traditional business culture relied heavily on physical approval stamps, known as hanko.
A few years ago, companies developed robots that automatically stamped documents.
From an engineering perspective, the solution was impressive.
The machine could:
- Detect document position
- Locate the correct approval location
- Apply the stamp accurately
- Process large volumes of paperwork
It was sophisticated engineering.
But I suspect Elon Musk would have asked a very simple question:
“Why are we stamping the document at all?”
The traditional workflow looked like this:
Document created.
Stamp applied.
Approval granted.
Japan’s solution became:
Document created.
Robot applies the stamp.
Approval granted.
The process became faster.
But the process itself remained unchanged.
Musk would likely approach it differently.
He would first ask whether the stamp serves any purpose.
If the answer is no, then the workflow becomes:
Document created.
Approval granted.
No robot required.
No automation project required.
No engineering effort required.
AI Risks Repeating the Same Mistake
This lesson feels even more relevant today.
Many organizations are rushing to deploy:
- Copilot
- ChatGPT
- AI Agents
- Workflow automation platforms
Sometimes those initiatives create real value.
Sometimes they simply become modern versions of the stamp robot.
Consider this workflow:
An employee writes a report.
AI summarizes it.
A manager reads the AI summary.
AI generates presentation slides.
AI creates meeting notes.
AI distributes action items.
Everything becomes highly automated.
Yet one question remains unanswered:
“Did anyone actually need the report?”
If the answer is no, then the organization has successfully automated waste.
Why Tesla Uses Weekly Cadence Reviews
The book mentioned Tesla’s practice of maintaining a regular weekly cadence for reviewing operations.
The interesting part is not the meeting itself.
The important part is the mindset.
Every week, teams ask:
“Why does this process exist?”
Most corporate meetings focus on adding new work.
Tesla appears to spend significant effort identifying work that can be removed.
That is a fundamentally different approach.
Can Large Companies Really Do This?
This raises an interesting question.
Can companies like IBM, Intel, Samsung, or major banks adopt this philosophy?
The answer is both yes and no.
It is difficult because large-company processes often emerged for legitimate reasons:
- Quality failures
- Customer requirements
- Regulatory compliance
- Internal controls
- Audit findings
The challenge is that while the original problem may disappear, the process often remains forever.
As a result, new procedures get added every year.
Very few are removed.
Large organizations accumulate processes the way old software accumulates legacy code.
The Lesson
What makes Musk’s framework powerful is not automation.
It is the order of operations:
- Is the task necessary?
- Can it be simplified?
- Should it be automated?
Most organizations start with Step 3.
The largest productivity gains usually come from Steps 1 and 2.
Conclusion
As AI becomes more powerful, automation becomes easier than ever.
Ironically, that makes Musk’s philosophy even more important.
The key question is no longer:
“What can we automate?”
The more important question is:
“What shouldn’t exist in the first place?”
The Japanese stamp robot may have been an impressive engineering achievement.
But from Musk’s perspective, it may also represent a classic example of automating a process that should have been eliminated.
Perhaps the most valuable innovators in the AI era will not be the people who build the smartest AI systems.
Perhaps they will be the people who recognize that some work never needed to be done at all.