Office Work

23 April 2018

Today I spent more than nine hours ‘at work’, though this was largely thanks to the innovation of distinguishing ‘at work time’ from time spent literally working. And especially from time spent literally productively working, on work.

This was recommended to me by several people, I think partly as a way to be paid a similar amount for similar efforts to other people — who apparently also do not work so much at work — and partly because somehow being vaguely oriented toward doing work, and then doing what seems good, even if it is a mixture of work and staring into space and eating rice pudding and walking around the block — often seems better for productivity than working intensely for however long you can, then stopping your timer and being ‘not at work’ the moment that your stamina fails.

Another popular work innovation is ‘go to an office that isn’t your bed’. And I did that too! Today was a first attempt at gathering various people working on small AI risk mitigation things into one place to work, for the purpose of enjoying that water-cooler conversation that I am told made Bell Labs great. And it was good! Even for its intended purpose, I think: I heard from multiple people about how I could maybe do my job better, and learned gossip about the other AI safety projects, and heard about a potential job candidate. And there was rationalist erotica recommendation, and dancing, and manual labor, and food, and feedback on how to correctly do squats.