Reading this as a machine learning person who spent NeurIPS asking authors “what do you want this work to make possible?” and getting a lot of “uhhh” in response, your secret third real finally gives me a name for why that felt so off.
The distinction between real 1 and real 2 and this third sense matches almost exactly the filter I’ve been groping toward. I keep running into papers where the math is fine and the benchmarks click, but I can’t tell what the authors think should happen in the world as a result, or what phenomenon they actually care about understanding. Your classification makes that discomfort legible instead of just grumpiness.
I also really appreciated the willingness to kill projects once you realised they weren’t going to be “real enough” in that third sense. That’s a standard I’d like to hold myself to more often. This essay is going to sit in the back of my head the next time I’m tempted to massage a negative result into something paper-shaped.
I love this post. I’m not in academia, but I feel this so deeply — the question of what makes work meaningful or truly valuable feels so important, especially when it feels like the world is taking shortcuts and incentives the wrong type of work
Would you say academia or industry rewards the third kind of real by your definition, or it feels more like a personally calling to the things you care about and any reward would be secondary to satisfying this calling.
thank you for writing this! i've been struggling with very similar thoughts this year and your vocabulary for realness has helped crystallize some of my very abstract thoughts/feelings about giving up writing papers for the sake of writing papers and the ick i get when i read papers that were clearly written for the purpose of being written. i feel like we're in the minority in the research community, which makes me sad, but reading posts like this one remind me that at least i'm not alone 🫶
Reading this as a machine learning person who spent NeurIPS asking authors “what do you want this work to make possible?” and getting a lot of “uhhh” in response, your secret third real finally gives me a name for why that felt so off.
The distinction between real 1 and real 2 and this third sense matches almost exactly the filter I’ve been groping toward. I keep running into papers where the math is fine and the benchmarks click, but I can’t tell what the authors think should happen in the world as a result, or what phenomenon they actually care about understanding. Your classification makes that discomfort legible instead of just grumpiness.
I also really appreciated the willingness to kill projects once you realised they weren’t going to be “real enough” in that third sense. That’s a standard I’d like to hold myself to more often. This essay is going to sit in the back of my head the next time I’m tempted to massage a negative result into something paper-shaped.
I have this exact feeling about machine learning papers all the time I'm glad you relate!!! thanks for reading : )
I love this post. I’m not in academia, but I feel this so deeply — the question of what makes work meaningful or truly valuable feels so important, especially when it feels like the world is taking shortcuts and incentives the wrong type of work
i enjoyed this post!
Would you say academia or industry rewards the third kind of real by your definition, or it feels more like a personally calling to the things you care about and any reward would be secondary to satisfying this calling.
I think real 3 works for some people in industry.... unfortunately I am allergic to having a job
mother has mothered again truly
thank you for writing this! i've been struggling with very similar thoughts this year and your vocabulary for realness has helped crystallize some of my very abstract thoughts/feelings about giving up writing papers for the sake of writing papers and the ick i get when i read papers that were clearly written for the purpose of being written. i feel like we're in the minority in the research community, which makes me sad, but reading posts like this one remind me that at least i'm not alone 🫶