Remove Amazon employment references from moratoriums post

Replace specific Amazon anecdote with generic explanation of change
freeze problems. Maintains the same key points about why moratoriums
fail without referencing any specific employer.

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

Author: cloudartisan

content/posts/2025-08-04-why-change-moratoriums-dont-work.md

@@ -17,13 +17,11 @@ But I've learned from experience that change moratoriums, while well-intentioned
 
 Here's what actually happens when you freeze deployments: all those changes that were ready to go don't just disappear. They pile up, waiting for the moratorium to lift. Instead of deploying incremental changes in isolation, you end up with massive, unwieldy releases that combine weeks or months of work.
 
-I witnessed this firsthand during my time at Amazon. We used to do release freezes for about six weeks over the holidays - only critical bugs could go out, and only after extensive review. Two things consistently happened:
+When deployments freeze, all those changes that were ready don't just disappear - they pile up. Instead of incremental changes in isolation, you end up with massive, unwieldy releases combining weeks or months of work. This creates a dangerous situation where:
 
-1. **We still broke production.** Even with minimal changes and extra scrutiny, incidents still occurred. The freeze didn't eliminate the fundamental issues.
+1. **Production still breaks.** Even with minimal changes and extra scrutiny, incidents still occur. The freeze doesn't eliminate fundamental issues.
 
-2. **Post-freeze releases were disasters.** I remember one year we had nearly 400 changes in a single release (we were used to about 15). That release was attempted and rolled back three times before it finally succeeded.
-
-Amazon eventually moved away from release freezes entirely, adopting continuous delivery instead. When I left, changes were flowing to production roughly 10 times per day, and production defects were trending downward.
+2. **Post-freeze releases become disasters.** Accumulated changes create complexity that's difficult to test and risky to deploy.
 
 ## Stability Through Stagnation Doesn't Work