The start of a fire is a criminal investigation, not a root cause analysis. Billions of dollars are on the line, because different groups have to pay out depending on how/who started the fire.

In this case there was no criminal liability. That means all the fire insurance companies will have to pay out of their own pocket.

For other fires, PG&E was found liable for not maintaining their power lines. They will probably have to declare bankruptcy, because the fire insurance companies will all sue them to cover their payouts.

There is a different team that does RCA and puts in place new policies or proposes new regulations to prevent whatever caused the fire, but what you're seeing is the result of a criminal investigation (although they do work together in collecting evidence).

I was reading the book "What Do You Care What Other People Think?" by Richard Feynman, where he describes being on the committee investigating the loss of the Challenger Space Shuttle.[1]

What I found interesting was his take on the software system:

"To summarize then, the computer software checking system and attitude is of the highest quality."

At first I thought, "wow, software is getting a pass."

But then mulling it over, I realized that software is almost easy compared to the unpredictability, unreliability and untameable stubbornness of the physical world.

[1] https://science.ksc.nasa.gov/shuttle/missions/51-l/docs/roge...

The rub here is that you can't just make nature and urban-forest interfaces 'wildfire-safe' because that means cutting down all the vegetation, replacing it with non-native species, or simply removing all access to the vegetated areas.

There's really nothing to do besides clear low brush every so often... which fire is better at than humans anyway.

Controlled burns are a thing for a reason.