Yeah, I don’t think the phrase “learning curve” has any built-in suggestion, even culturally, to imply that the reasonable default assumption is one way or the other. I only ever heard learning curve to refer to something getting easier after awhile, which is indeed a valid curve
Yeah this is a common misunderstanding I’ve had to clarify to people as well, even people who work in tech. I support only using “Law” for things that are scientifically actually laws. I don’t even like to use it as a joke (Murphy’s Law) because, unbelievably, some people really do take that to be a law of the universe too.
Yeah, a lot of these things actually do make sense, just in a more precise way than even the people using them intend. Gravitational pull is also like this. Earth’s gravitational pull is not weak, it literally keeps everything on Earth tethered to it. More importantly, it happens as an intrinsic property of the Earth, the Earth doesn’t need to “try” to exert gravitational pull on things. Furthermore, gravitational pull attracts more mass which begets even more gravitational pull, like a snowball effect.
So gravitational pull is not about the strength of the force, but the fact that it is natural, effortless, and often forms a positive feedback loop (borrowing from another comment here lol).
So if I say someone at work has a lot of gravitational pull, I’m conveying that they do a good job of bringing other people into their area or work, that they naturally do it almost without even trying to, and that as their social influence grows, they just end up with even more social influence. It’s a really deep metaphor which is also physically accurate.
Hm, this is interesting. I only have a passing understanding of control theory, but couldn’t a positive feedback loop indeed be good when the output is always desirable in increased quantities? A positive feedback loop doesn’t necessarily lead to instability, like you said. So maybe this is just me actually-ing your actually, lol.
As for “more optimal”, oof, I say that a lot so maybe I’m biased. When I say that I’m thinking like a percentage. If optimal is X, then 80% of X is indeed more of the optimal amount than 20% of X. Yes, optimality is a point, but “more optimal” just seems like shorthand for “closer to optimal”. Or maybe I should just start saying that?
This reminds me of a professor I had who hates when people say something is “growing exponentially”, since he argued the exponent could be 1, or fractional, or negative. It’s a technically correct distinction, but the thing is that people who use that term to describe something growing like x^2, are not even wrong that it’s exponential. I feel like when it comes to this type of phrasing, it’s fine not to deal with edge cases, because being specific actually makes what is said more confusing.
“I’m in a negative feedback loop with respect to my laziness which will soon stabilize with me continually going to the gym daily, which is closer to optimal than before. As a result, my energy levels are going to increase exponentially, where the value of the exponent is greater than 1!”
Hmm. Now that I say it that doesn’t seem that crazy. Although I do still think some common “default settings” don’t do any harm.
The problem is the location of the steepness makes the difference between whether this means it’s easy first and slow progress later, or slow progress first and easy later. Is it like, x^1.5, or is it like ln(x)? Both are very steep at some point.