2 ms·
The equation should have per-subject weights to modulate the "how many people" part. For instance, if my app helps Mr. Musk be 10% more productive, that's more
by fchollet 10y ago
The equation should have per-subject weights to modulate the "how many people" part. For instance, if my app helps Mr. Musk be 10% more productive, that's more impactful than helping Mr. Smith be 10% more productive.
- tscizzle 10y agoI assume the equation he's talking about already propagates the value to the "leaves", where if the app is boosting a person's productivity, and that person's productivity increases value for a group of people, then it's that final group of people's increased value that is counted toward the original total. And there could be more than 2 layers before you quantify the value without recursing any deeper, of course.
- metaphorm 10y agoabsurd hero worship. I like Elon Musk just fine, but you're fawning over him. How do you propose we nominate "great persons" for your "equation" anyway?
- i2shar 10y agoThe weights would be different based on who's assigning them.
- Whitestrake 10y agoHypothetically, this is unnecessary. Assuming that a 10% increase in productivity causes a 10% increase in the impact on people (either in quality or quantity), 10% of Mr. Musk's impact may be greater in absolute terms than 10% of Mr. Smith's. One may debate whether 10% productivity = 10% impact, but if that increase is attributed back to you, there's no need to artificially manipulate the equation.