> If you’re like most of us on HN, at work, when we make big mistakes that cost our companies money and users, we have thorough retrospectives, learn from our mistakes, identify root causes, build for redundancy and high availability, and prevent the same incident from ever happening twice.
Depends on the process. If the retrospective process is a big PITA that goes nowhere, and management is only going to support changes that it wanted before the incident, regardless of any knowledge uncovered by the evidence, then minimize the reporting and hide under a rock for a while, try not to get management attention again.
Which scenario seems more applicable to a government lab in Wuhan?
I don't think we're going to get the straight information from the lab, because the incentives aren't there. If they admit they did it, everybody is mad, and CCP has to take action to save face, and the lab team doesn't want to face that. If they admit they didn't do it, nobody believes them.
Outside of finding some lucky samples, like a good match from outside the area in say October or November, I don't think we're going to get anything definitive from forensic analysis of virus DNA. I'm not an expert in the field, but the same sequence gets interpreted different ways by different experts; and usually with some wishy-washyness.
The data quality for the early pandemic isn't getting any better or more independent from the same officials we don't trust.
If we can't get a story we trust from the horse's mouth, and we can't figure it out through forensic analysis, the best way forward is to figure out what interventions would be reasonable if we assume lab origin and what would be reasonable if we assume natural origin and do the practical interventions.