5 ms·

Publish or perish is yet another example of Goodhart's Law (https://en.wikipedia.org/wiki/Goodhart's_law) https://en.wikipedia.org/wiki/Goodhart's_law). Paper

Publish or perish is yet another example of Goodhart's Law (https://en.wikipedia.org/wiki/Goodhart's_law) https://en.wikipedia.org/wiki/Goodhart's_law). Paper count is a terribly easy metric to game when used to evaluate academic output because publishing low quality papers doesn't count against the authors. We can fix the problem in an unbiased manner by using ChatGPT to quantify article novelty, impact, and quality. /s

_aavaa_ 2y ago
I almost missed the /s in my anger while reading that last line. It's too close to being real...
- ben_w 2y ago
  You might want to get a paper bag to scream into: https://arxiv.org/abs/2408.06292 https://arxiv.org/abs/2408.06292
  - _aavaa_ 2y ago
    They say AI scientist, but it looked more like the NN Evolution by LLM Selection.
OvbiousError 2y ago
You also need citations though, and those don't come with low quality papers, nor will they help your reputation. To get citations people have to actually read your stuff.
- inglor_cz 2y ago
  Or organize a fraudulent citation ring.
  - dguest 2y ago
    Some scientific collaborations put thousands of people on every paper, publish around 100 papers a year, and cite at least 10 of their own papers in each one. To some, that might be "a fraudulent citation ring". To others it's just business as usual. Of course a lot of academia has caught on to the ruse, but the mitigations are pretty ugly: some funding agencies and universities just stopped counting papers from large experiments as valid scientific contributions.