General > General Technical Chat

Diversity, Equity and Inclusion

<< < (21/76) > >>

emece67:
I realize now that my opinion about some other people on this thread has changed. In some cases dramatically and to worst.

There is the obvious possibility that the opinion that some other members of the forum have about me to also have changed in the same sense.

This is why I think that it is a big error to allow this kind of threads.

For me it is over.

Enjoy.

james_s:

--- Quote from: emece67 on August 18, 2022, 07:36:56 pm ---I realize now that my opinion about some other people on this thread has changed. In some cases dramatically and to worst.

There is the obvious possibility that the opinion that some other members of the forum have about me to also have changed in the same sense.

This is why I think that it is a big error to allow this kind of threads.

For me it is over.

Enjoy.

--- End quote ---

And that illustrates precisely what people are complaining about. You have come right out and said that you have drastically changed your opinion of some people after finding out that they have a different perspective than you on some topics and because you are offended by somebody saying something you disagree with you are advocating that they be silenced. Yes there is a possibility that some are so fragile that exposure to a differing view could drastically change their opinion of others, but most of us will carry on and continue to engage in friendly discussions on other topics with them. You have not presented a counter point at all, instead you are pushing to have the discussion shut down. That is a nearly perfect example of the issue we are discussing, a vocal minority of people when faced with something they disagree with choose instead of presenting points in favor of an alternate point of view, to attempt to simply shut down the opposing view. I don't know, but I suspect that there is some cognitive dissonance going on, they lack intellectual points to support their view and instead it is based on how they feel. I admit I have never really understood why some people take some idea so personally that they themselves feel attacked when somebody points out flaws with that idea. An extreme example of this behavior are the cases where people have been brutally attacked for rooting for the "wrong" sports team.

KaneTW:
It's a defensive reaction. Not positing that this is the case here, but:
* Disagreements often provoke the same physiological response as a physical attack
* When your worldview is being challenged, you instinctively want to defend it. When you don't have arguments to defend it with, you lash out.

For a lot of people, specific issues are a central tenet of their identity. Challenging it is like waving a red flag at a bull, but unfortunately we can't just gore people we don't like over the internet.


I see this with gender identity, Ukraine/Russia, etc, etc. Any contentious highly emotional issue, basically.

Echo chambers just make it worse (Twitter, State TV, ...)

Zero999:

--- Quote from: HuronKing on August 17, 2022, 05:00:03 pm ---Here's something else to ponder. Companies that have used AI to find candidates, even in the efforts of perfect equality, often just end up perpetuating the racist stereotypes that built out the makeup of their workforce in the first place (because the AI is trained on data of a workforce that was historically built on discrimination by race, gender, or sexual orientation):
https://en.wikipedia.org/wiki/Artificial_intelligence_in_hiring#Controversies

And of course, this:
https://xkcd.com/385/

--- End quote ---
AI only relies on the data presented to it. It's possible to build an algorithm which performs well at predicting something, yet also yields politically incorrect results. For example, you could interrogate a database of postal addresses to find out who's most likely to default on a loan, or make an insurance claim and you'll find they're most likely to be the more deprived areas, with more of a certain ethnic minority group. The algorithm will be branded as racist, but it's working properly and generating good results. The problem is with society not the algorithm and the same applies to AI. Unfortunately it's not possible to fix society so everyone is equal, without resorting to unfair or even authoritarian social engineering, so we're stuck to some extent.


--- Quote from: KaneTW on August 18, 2022, 08:39:34 pm ---It's a defensive reaction. Not positing that this is the case here, but:
* Disagreements often provoke the same physiological response as a physical attack
* When your worldview is being challenged, you instinctively want to defend it. When you don't have arguments to defend it with, you lash out.

For a lot of people, specific issues are a central tenet of their identity. Challenging it is like waving a red flag at a bull, but unfortunately we can't just gods people we don't like over the internet.


I see this with gender identity, Ukraine/Russia, etc, etc. Any contentious highly emotional issue, basically.

Echo chambers just make it worse (Twitter, State TV, ...)

--- End quote ---
I agree. I think many of use have been there, myself included. I admit, it's difficult to  look back on the situation or debate and come to the conclusion I was wrong. It's something I'm working on and getting better at doing. Part of the problem is, when someone disagrees, they feel compelled to take more time offering an explanation why they think the other person is wrong, than when they simply agree with them. This is of course human nature, since preaching to the converted requires less effort.

It is frustrating how many people seem to have difficulty with disagreeing with someone one one thing, yet finding common ground on another. To demonstrate this is possible, I've picked out a post which I partially agree with.


--- Quote from: sokoloff on August 17, 2022, 04:01:57 pm ---In engineering (software engineering in my specific case), I think there is an observable outcome that suggests a strong possibility that some effects are happening to cause the profile of working engineers to be very different from the profile of the overall population.

Now, that could happen from a bunch of reasons that might be completely okay, supportable, understandable, or even beneficial.
Or it might be from biases or filters that we should work to reduce.

For my own actions, I'm very supportive of examining the questions and working to reduce apparent biases (or even introducing offsetting biases) when they involve groups of people.
As a concrete example: I fully support adding all-women's or historically-black colleges with strong tech programs to the set of campuses that we recruit at. I also support something like the "Rooney Rule" for professional hiring levels.
--- End quote ---
I strongly disagree with you here. I deem any policy which treats people differently, based on irrelevant characteristics, i.e. interviewing quotas for minorities, is a good idea. My dad had that problem when he worked at the council. HR commented on how he didn't select any CVs of ethnic minorities and wanted him to interview a group of minority candidates, who were clearly unsuitable. It wouldn't have been fair for them to have wasted their time travelling to the interview, especially given they were from further afield. What was even more silly is most of the information giving away their ethnicity had already been removed form their CVs by HR.

Everyone has their biases and there's little evidence to suggest they can be changed, via training. I admit it's not a bad thing to make people aware of any biases they may have, but they can't be improved or eliminated.


--- Quote ---I do not support introducing "offsetting biases" when it comes down to individual people.
As a concrete example: While I fully support recruiting at Wellesley College or Florian A&M, I don't support having a different hiring bar for candidates from those schools versus others. Similarly, while I support the Rooney Rule (or equivalent), I don't support differences in promotion or hiring criteria to give "diverse candidates" an advantage. I find the desire to award such advantages understandable but still significantly undesirable.

IMO, you fix inappropriate discrimination with a strong move towards fairness and impartiality, not with counterbalancing/offsetting discrimination.

Inappropriate discrimination is unfair, undesirable, and value-destroying regardless of the motivations of the people practicing that discrimination.

--- End quote ---
I agree everyone should have to meet the same criteria, when applying for a position. As you said in another post, lowering the bar is bad for those it's supposed to help.

tom66:

--- Quote from: Zero999 on August 18, 2022, 09:28:43 pm ---AI only relies on the data presented to it. It's possible to build an algorithm which performs well at predicting something, yet also yields politically incorrect results. For example, you could interrogate a database of postal addresses to find out who's most likely to default on a lone, or make an insurance claim and you'll find they're most likely to be the more deprived areas, with more of a certain ethnic minority group. The algorithm will be branded as racist, but it's working properly and generating good results. The problem is with society not the algorithm and the same applies to AI. Unfortunately it's not possible to fix society so everyone is equal, without resorting to unfair or even authoritarian social engineering, so we're stuck to some extent.
--- End quote ---

I'm reminded of car insurance companies who used to be able to charge women less - the idea was that women, in general, have fewer accidents and the accidents they do have tend to be less serious.  Men, in contrast, are more likely to be aggressive when driving, more likely to drink-drive, and more likely to drive when tired.  And, they're quite likely to do any one of those when they have others in the car with them.  So the average payout is higher.

But, it's not legal any more to charge less to women (or more to men) on the basis on gender alone.  They can charge more or less based on insurance history, so an average female driver will pay less than an average male driver.  But for new drivers, or those with limited history, the price has to be the same.

It's completely bonkers - why shouldn't you be able to charge on actual risk?  Why do we pretend these things are equal when they quite clearly are not?

Navigation

[0] Message Index

[#] Next page

[*] Previous page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod