“Fact-Checking” Takes Another Beating

I feel like the biggest issue facing journalism is the lack of any kind of engineering mindset in the field. Even though journalism theoretically has some clear goals to work towards (more accurate, less biased, faster, clearer etc), as a field it doesn't progress.

Although I enjoy his articles, Taibbi is symptomatic of this trend. He writes a lot about the problems of journalism but never spends time reflecting on how to engineer better outcomes. It's all just random people writing stuff, and occasionally (apparently) these quasi-mythical fact checkers who are just other random people double checking dates and definitions against Wikipedia. But there are no institutional processes worth anything and the response to failure is either obfuscation or occasionally an apology, but never a post-mortem with action items for new processes to prevent them from happening again.

There's a serious need for someone to engineer news. To say, here's where something went wrong, here's the post-mortem with a root cause analysis, here are the entries in our ticket system tracking the resulting action items and (ideally) software changes that will automatically check for these problems in future.