4 ms·
I imagine that to someone of your political persuasion, this is self evident. But humor someone who doesn't share your views and actually demonstrate (empirical
by omonra 11y ago
I imagine that to someone of your political persuasion, this is self evident. But humor someone who doesn't share your views and actually demonstrate (empirically for example) that you are right.
I personally don't care whether the geeks who work at a given tech firm are of particular gender or race, as long as they produce quality stuff.
I would care if someone couldn't get into the industry because they are X - but if these individuals don't want to - it's not something I consider a problem.
- AdeptusAquinas 11y agoNot sure where my political orientation comes into it. Also note I am not from the US, so it is unlikely I fit into whatever weird party conceptions you have over there anyway :) The point I was making that increasing numbers and diversity in something means more stuff is produced and there is greater innovation. It means the pool of candidates for a given role is larger, allowing better selection. This is all self-evident - as Bill Gates has said, doubling the size of your workforce cannot help but improve things.
- pgeorgi 11y ago> as Bill Gates has said, doubling the size of your workforce cannot help but improve things Especially for those who are paying the wages and don't like to be in a seller's market.
- omonra 11y agoOk - you think so, I don't. So we can agree to have a difference of opinion. However - on my side is every single real-life example of success in tech. EVERY SINGLE successful technology company was NOT diverse (and only once they get big enough to be targeted by diversity scammers that they start to pay lip service to this). Therefore if diversity was indeed at all important, wouldn't you be able to point to examples of more diverse companies being actually more successful? Heretofore the empirical argument is 100% against you.
- AdeptusAquinas 11y agoMy argument is that correlation doesn't imply causation. Just because all current examples are biased does not mean that the only way to do things is to be biased, right? To put it another way, the concept of disruptive change is when some company does something different than all the other companies have done up to that point. I imagine uber, airbnb et al being told by someone that the 'empirical evidence is 100% against them' as well.