← Back to context

Comment by throwaway0a5e

4 years ago

So why not save everyone the week of back and fourth emails while nothing gets done and ask them directly how they want to deal with it rather than putting tons of people on blast and substantially constraining their options by bringing intra-organization politics into the mix?

Sounds like a great way to get Normalization of Deviance[1]. One senior person says "I know how to dispose of this, so it's OK if I don't go through proper channels." Then the next person, following their lead without understanding the implications, says "Joe Senior over there disposed of something scary they found without wasting time going through EHS, so I'll do the same." Maybe it goes fine for a while, but eventually you'll end up with a situation where you've poisoned the groundwater or released dangerous chemicals into the air, because nobody is following the proper channels any more.

1. https://en.wikipedia.org/wiki/Normalization_of_deviance

  • Telling your boss or relevant colleague instead of going over everyone's heads from the get go isn't normalization of deviance and we both know it.

    I really dislike these sorts of "name drop" comments. They're just equivalent of "F" or "the front fell off" with a high enough brow for HN veneer on top.

    • You're suggesting that people bypass official procedures and/or laws in order to save time. This is a bad path to start down. The fact that you're posting this as a throwaway indicates that you don't want your HN account associated with these proposals.

      Here's a relevant software-related analogy:

      I work in a situation where if we receive certain types of data, we have to go through proper procedures (including an official incident response team). It would be very easy for me to say "I've verified that nobody accessed this data, and we can just delete it," instead of going through the proper channels, which are VERY annoying and require a bunch of paperwork, possibly meetings, etc.

      Maybe nothing bad happens. But next time this happens, one of my junior colleagues remembers that the 'correct' thing to do was what I did (clean it up myself after verifying nobody accessed the data). Except they screwed up and didn't verify that nobody had accessed the data in question - and now we are in legal hot water over a data privacy breach.

      And then people go back through the records, and both the junior engineer and I get fired for bypassing the procedures which we've been trained on, all because I wanted to save some time.

      1 reply →

It sounds like you may have had a bad time with EHS in the past. I found that by making friends with everybody involved ahead of time, I suddenly had excellent service.

sadly, after 30 years of training to be a superhacker on ML, my greatest value is actually in dealing with intra-organizational politics.

  • I work in a regulated software space, and my experience is that treating quality and regulatory folks as adversaries is a great way to have your projects take way longer than they should and cause immense frustration. Understanding the hows and whys of the way things work makes life easier for everyone. I haven't worked with EHS in the past, but I imagine it's much the same - if you're seen as somebody who's trying to cut corners and take shortcuts, yeah, you'll probably have a bad time.

  • Great EH&S people are amazing.

    At the root of researchers' reactions to EH&S are two things:

    1) EH&S will frequently be the source of unfunded and unresourced mandates. "You must stop your research until you've properly tidied up all of your electrical cords" is, in the short term, an impediment to forward progress. Researchers frequently under-budget for safety/disposal costs when submitting proposals, leaving nobody with money to foot the bill for expediting the resolution for a safety stoppage.

    2) The statistical likelihood of a single accident is greater for a large organization than for a single research group. One group can get away with a lethal practice for a hundred years before the first death. A hundred labs will only get away with a similar practice for a year.

    If you find a competent and reasonable EH&S auditor, they are a great resource. They may ding you for some safety violations, but they'll be able to point the way toward safer practices. In the best case, even if they don't have mitigation funds for longstanding problems, their voices carry real weight and can expedite the allocation of scarce resources toward real safety concerns.