I audited some embryology, neuropsychology classes in college. Still seems like a miracle, how the kids arrive...
(I won't argue about the magic moment of person-hood, if any. Life in squishy bodies on Planet Earth, we do our best. I'm glad for the chance.)
1. I think the dog has a cute name
2. I think the dog is cute
3. The dog doesn't attack me
4. Other dogs also attack people
5. It's the owner's fault for lack of training.
These infuriate me more than anything because they are all invalid. You also can't point this out because it's an emotional subject for them
It really is. It is very bizarre when I see people absolute flip out because someone points out that a dog that was bred to be violent ... actually becomes violent. And of course I know it isn't just pits that will attack (I'm looking at you chihuahuas!) but they combine that instinct with power and size and now you have a real problem.
Most dogs spend almost their entire lives docile. Those bread to be violent are ones that have high odds of a moment or two of uncontrollable violence.
(And if you look, it's almost always because they have a chihuahua or pinscher ancestor.)
The maternal womb is also considered an environmental factor, so by the time they are born the offspring has already been affected, greatly, by the environment. And that's precisely one of the major hypotheses of this paper.
Of course, genetics also influences our personalities greatly, but as seen in this paper with genetically identical fish; even in this case there is shown to be individuality. Now, this paper doesn't say anything about how much of one's personality is genetics or how much is influenced by the environment. What it says is that in genetically identical individuals that grow in (pretty much) identical conditions there is already individuality that's present at the time of birth which is possibly mediated by the pre-natal environment.
I also look at my own life and imagine I was born in a different country with different hardships. I can only imagine that some of my personality traits would have expressed themselves differently based on different survival necessities.
As a tech analogy, varying your training data will vary your outputs, even if the model parameters remain consistent.