6 ms·
Somewhere between 30-40 (rough estimate) percent of the people I know support Damore. Hardly a majority, but also an indicator that these are not the ideas of
by gavanwoolery 9y ago
Somewhere between 30-40 (rough estimate) percent of the people I know support Damore. Hardly a majority, but also an indicator that these are not the ideas of a lone heretic. I am not as concerned with the fact that Google fired an employee for having an opinion, but more concerned with the fact that they only fire people with opinions that do not match those of the majority. Even if Damore's opinion's were wrong (which, according to several scientists, they are not), it should be ok to pose a theory without subjecting yourself to a potential witch hunt.
- throwaway_2343 9y ago> it should be ok to pose a theory without subjecting yourself to a potential witch hunt. That theory is that a large portion of your coworkers are unfit for the job because science. And they were only hired because of misguided politicking.
- deepsun 9y ago> they only fire people with opinions that do not match those of the majority. No, they only fire people that boiled media. For Sundar, it doesn't matter what the guy actually wrote, what matters is what media wrote about it. And they pictured him as a misogynist. So that is the new truth, now deal with it.
- Caveman_Coder 9y ago> "...that is the new truth, now deal with it." I'd want my employer to have a spine and defend me, not kowtow to the flash-in-the-pan media opinions...
- dclowd9901 9y ago> which, according to several scientists, they are not Good thing that's enough for a quorum. /s There is absolutely no evidence that there is a biological imperative that prevents women from being as effective as men at software development. None. Zero. Zilch. Just about every disparity you can imagine can be categorically dismissed by upbringing and cultural side effects. It doesn't even pass the sniff test: do you really think there's something inherent to the Y chromosome that allows better rote analysis?
- beaner 9y ago> There is absolutely no evidence that there is a biological imperative that prevents women from being as effective as men at software development. The memo doesn't make this claim.
- dclowd9901 9y agoThen in a pure meritocracy, why would they not succeed? Ergo the memo itself is a straw man.
- beaner 9y agoDid you read the thing? It was very clearly about statistics. The message of the memo was not "women are biologically unfit for tech," it was, as a biological tendency, women are more likely to have traits which disincline them to have an interest in tech-related things, and therefore end up in tech. To me this says very obviously that women can and are intellectually superior to men on individual cases. But we would not expect an even 50/50 distribution of tech interest among genders. Feel free to refute that as a separate claim, but if your take from the memo was "all women are biologically inferior and cannot match or be better than men," then I would say you were not being intellectually honest when reading it.
- zbobet2012 9y agoIt doesn't make that argument either. It makes the argument that since less women (will on aggregate) be interested in CS then men, less women will be in CS. And that less women being interested in CS has _some_ basis in biological differences. Whether you agree with that or not (the research does support it pretty strongly, but is certainly not settled), it is a reasonable argument.
- patrickyan 9y agoThe memo literally lists the biological/psychological traits of men and women that may have caused the gender gap in tech. (He tries to clarify that he is talking on a population level and not talking about individuals)
- peoplewindow 9y agoI think that's what a lot of people are most concerned about: not the gender science but the new information about Google's culture. Google isn't a normal company of normal importance. It runs the dominant search services in the western world on which over a billion people rely on for information. Bing hardly comes close. Old media hasn't picked up on it yet, but a series of Googlers are giving anonymised interviews. They are making a lot of very disturbing allegations: Google is manipulating recommended content on YouTube to suppress conservative videos and viewpoints. That the voices pushing to manipulate web search are very loud. That Google Research ran a programme to investigate why people who get very good interview scores do worse in their career than people who get more mixed feedback, discovered it was due to bias in hiring towards Ivy Leaguers and under-represented minorities that caused them to have a lower bar. That YouTube systematically punishes channels that cover right wing politics. Senior management being on the verge of tears after Trump won. Conservatives getting punched. Replies to people like "isn't it nice to be white" or "congratulations on your white penis". People sabotaging each others performance reviews for not being on board with social justice wars. Google makes a lot of money because people search there and because they have little incentive to go anywhere else. Bing and other well funded search engines aren't dramatically worse than Google, they just aren't better, so they can't build up a userbase. If people stop trusting Google's results, if they start believing they're manipulated and unreliable, suddenly they have an incentive to check out the competition. That would lead to more ad clicks, more money to invest in the search engine, better results, etc. So that seems like a problem for both Google and its users. If you aren't allergic to Breitbart you can read the interviews here: http://www.breitbart.com/tag/rebels-of-google/ http://www.breitbart.com/tag/rebels-of-google/
- lancewiggs 9y agoGoogle is a global company. What might pass for "right wing" in the US is lunatic fringe in many other countries. Do yes the search engine of the works should be biased towards fact based sources and away from sites like BB that preach hate and not science.