Courage Is Not the Reason

Courage Is Not the Reason | Generated by AI

Home 2026.07

Question: Why do people have the courage to start a startup?

Answer:

Most founders do not actually feel “courageous” before starting. They usually start because the expected value of trying becomes higher than the expected value of staying.

A startup decision is roughly:

[ EV = P(success) \times Reward - P(failure) \times Cost ]

For example:

Traditional job:
    salary: $100k/year
    probability of major upside: low
    downside: stable

Startup:
    probability of success: 5%
    outcome: $10M company
    failure cost: 2 years + opportunity cost

Expected value:
    0.05 * $10M - cost
    = potentially attractive

The founder is not thinking “I am brave”. They are often thinking:

“If I don’t try this, I will regret it.”

1. They see an asymmetric opportunity

Many founders have this mental model:

Worst case:
    lose 2-3 years
    learn a lot
    return to employment

Best case:
    create something valuable
    gain freedom
    change their life

The downside is limited, but the upside is huge.

This is especially true for engineers because skills are portable:

coding ability
+
AI knowledge
+
system design
+
network

= can recover

A 25-year-old engineer quitting a job has a different risk profile than someone with no transferable skills.

2. They are already frustrated with the existing world

Many startups begin from:

"I hate this problem"
        +
"I know a better way"
        +
"Nobody is fixing it"

Examples:

Airbnb founders: hotels were expensive, spare rooms existed.
Uber founders: transportation was inefficient.
OpenAI founders: AI development needed a new organization model.

The emotional driver is often not courage; it is obsession.

3. They reduce risk before jumping

Good founders rarely jump blindly.

They create a “minimum viable escape”:

Night/weekend:
    build prototype

↓

users appear

↓

revenue appears

↓

quit job

A lot of successful startups started this way.

For an AI engineer today:

Month 1:
    build agent tool

Month 2:
    100 users

Month 3:
    $2k MRR

Month 6:
    decide whether full-time

The internet makes this possible.

4. They have a different relationship with failure

Many people think:

failure = identity damage

Founders often think:

failure = data

Example:

Attempt 1:
    nobody wants product

Learn:
    wrong customer

Attempt 2:
    different market

Learn:
    pricing wrong

Attempt 3:
    works

They treat startups like reinforcement learning:

environment
     |
     v
 action
     |
 reward/error
     |
 update policy

5. Timing matters

People become brave when the environment changes.

For example, AI has changed the equation:

Before:

1 engineer
+
idea

cannot compete with large companies

Now:

1 engineer
+
Claude/GPT/Qwen
+
open source models
+
cloud GPU

can build a serious product

The cost of experimentation dropped dramatically.

6. Many founders are already “forced”

Some start because:

cannot find a good job
dislike corporate politics
want autonomy
have financial pressure
see a market gap

Necessity creates courage.

For someone with your background (shipping products, training models, building agents), the interesting question is not “do I have courage?”

The more technical question is:

Can I find a 10x problem
where:
    my AI advantage is real
    users pay
    distribution is possible?

The modern founder loop is:

observe problem
      ↓
build agent/product in days
      ↓
put in front of users
      ↓
measure
      ↓
iterate

The barrier is less courage now. The barrier is usually finding a painful enough problem and moving fast enough.

Back Donate