3 ms·
No, that’s not a fallacy. Your rebuttal is a bit of a fallacy because it talks about what is a “good” thing. I am not talking about good and evil or right and w
by lawnchair_larry 6y ago
No, that’s not a fallacy. Your rebuttal is a bit of a fallacy because it talks about what is a “good” thing. I am not talking about good and evil or right and wrong. I am talking amorally about cause and effect in complex systems. Making an analogy to slavery to illustrate a moral wrong isn’t responsive to that point.
Different systems produce different results. It tends to be the case that when you drastically alter the fundamental properties of a system, you won’t get the same results - because it isn’t the same system.
If you want a system where a company is not the property of its shareholders, and you fail to observe that all current implementations of that idea have never demonstrated a single instance of producing Google-like companies since the industrial revolution, your ideas are not going to be very robust.
If you don’t want that, sure, this is a perfectly valid position. But you are not to be taken seriously if you make these claims while earning 7 figures at Google. You haven’t even demonstrated that your ideas are viable to you, let alone all of society.
And since the principles of capitalism apparently quite unironically elude you, yes, they expect their own bank account to be padded more than yours (as it must, game theory prevents the transaction otherwise), which is why they fired her.
And the conditions that make Google better for them were not created exclusively by employees and especially not by employees who didn’t share the mission. Nor were they created and sustained without leadership. Nor were they created without venture capital. If you cherry pick certain properties of Google and expect to be reproduce those results, you aren’t making payroll. It is only through the right balance of incentives that such a system can function at all.
- pron 6y ago> Different systems produce different results. It tends to be the case that when you drastically alter the fundamental properties of a system, you won’t get the same results - because it isn’t the same system. Yes, but you claim to know changing which set of variables, by which amount, produced what result, while you don't actually know that. Your epistemology is emotional rather than rational. > and you fail to observe that all current implementations of that idea have never demonstrated a single instance of producing Google-like companies since the industrial revolution, your ideas are not going to be very robust. Not that I think Google is a great company (it's a pretty lousy one, all things considered) but it is unique, and still you don't know precisely what combination of variables unique to the US produced this particular result. > which is why they fired her. That's fine, but you blamed them for being sort of victims in the first place. > And the conditions that make Google better for them were not created exclusively by employees and especially not by employees who didn’t share the mission. I never said that. > It is only through the right balance of incentives that such a system can function at all. Sure, but there can be more than one right balance, and you don't know all of them. You make the mistake of thinking that if the current situation is X and you think X is good, then any other balance Y is likely worse than X.