I am no fan of Trump, but this is exactly why Google should be broken up. Google, Facebook, Twitter, etc.'s control of the digital public square is a threat to discourse and free speech. And Google is the worst offender of them, given how monolithic and ideologically skewed their internal corporate culture is.
I'll try to practice some charitable interpretation[0] and point out that she might not mean this in the sense of "small companies can't undermine democracy." Rather, I see this making sense if we consider it to be about how companies providing platforms and tools for mass information will always exist, as it makes economic sense for them to exist. But at the same time, if these companies are small, their platforms are prone to be attacked and abused by adversaries. And they need to have a certain size to stand a change against there adversaries, to be able to protect the people from them.
The quick cut editing between her and an anonymous “insider” seems questionable. When she says “next Trump situation” perhaps she is speaking of all the Russian bots and trolls. Otherwise, why didn’t she say “prevent the next Trump” but used the word “situation”?
2020 is likely to be worse with the improvements to Deepfakes. Most people can barely detect false stuff now and constantly reshre bullshit on Facebook that takes all of 5 seconds to debunk, just imagine GAN and GPT-2/BERT powered trolling.
I can remember trying to recall details of political stories. I type the breadcrumbs into Google and am unable to find what I'm looking for. I then realized that I have better luck on certain subjects when I try duckduckgo, and I know it's not because they have a more mature search algorithm.
If anything, I implied that DDG was worse in general, but only better when it came to certain political subjects. I take your meaning as rhetorical, but I'll answer anyway.
Several months ago, I read an article on Reddit about how the USPS was raising prices. In the comments, the discussion was of why it was cheaper to ship some goods from China than it was to ship from state to state. I had recalled something about Trump taking some measure against this, but apparently at that time nothing had changed. I tried finding more info on Google, but I couldn't find what I was looking for. On DDG, I located an article from Newegg describing the UPU, and how these deals were established many decades ago and gave 3rd world countries at the time, such as China, a break. I then found another article from thehill.com that explained that Trump had begun the one-year notice of withdrawing from the UPU, but until then the rates would be the same.
On another occasion, I was in a thread that was talking about middle class wages, and I noticed that on Google, I could find many articles about how the middle class was shrinking. On DDG, I found some of those articles, but also found information on how if you control for inflation and hours worked, middle class wages actually increased since the 90s.
>"We're also training our algorithms... if 2016 happened again, would we have, would the outcome be different."
>"2020, certainly on top of my old organization, Trust and Safety, top of mind, they've been working on it since 2016, to make sure we're ready for 2020"
>"My definition of fairness and bias specifically talks about historically marginalized communities, and that's who I care about. Communities who are in power and have traditionally been in power are not who I'm solving fairness for."
>"Our definition of fairness is one of those things that we thought would be like, obvious, and everyone would agree to and it wasn't. There was, the same people who voted for the current president who do not agree with our definition of fairness."
Her definition of fairness is an ideological definition. It doesn't mean "fair," it means weighted, that is not objective, that is manipulated towards her ideological goals.
>"For example, imagine that a google image query for "CEOs" shows predominantly men... even if it were a factually accurate representation of the world, it would be algorithmic unfairness."
>"ML fairness... So they're trying to modify the model such that even if the data for female CEO says... its low... It still balances out."
There is really no doubt about the political ideology driving all of this, but what is apparent more than all of that is that they believe they have a duty to manipulate reality as much as possible and as needed, in order to fit their user's perception with their ideological narrative. And it is basically abusing their power (with ML) to enforce a delusion.
>"Alignment with the narrative."
>"People (like us) are programmed."
This is Ministry of Truth levels of social engineering/programming. They don't want to have to defend their ideology, they want to eliminate all opposition, and I don't know how any free thinking individual can defend that, much less support it. Its totalitarian, tyrannical, and the most terrifying aspect of it all is that they are knowingly committed to the construction of an entirely fictional, delusional narrative version of reality.
So much bias, had this been liberal news media coming out with quotes from Google employee on how they plan to stop Bernie or Warren, then shit would have hit the fan. This post would have had 1000's of comments on why Google should be broken up.
This thread got flagged off of the front page within 20 minutes of being posted. It had a crazy velocity, then suddenly, gone. I guess some people aren't interested in relevant tech news if it hurts their narrative.
19 comments
[ 3.3 ms ] story [ 43.8 ms ] threadWhat are Google’s offenses in this domain?
given how monolithic and ideologically skewed their internal corporate culture is.
How are you measuring that?
[0]: https://en.wikipedia.org/wiki/Principle_of_charity
2020 is likely to be worse with the improvements to Deepfakes. Most people can barely detect false stuff now and constantly reshre bullshit on Facebook that takes all of 5 seconds to debunk, just imagine GAN and GPT-2/BERT powered trolling.
Several months ago, I read an article on Reddit about how the USPS was raising prices. In the comments, the discussion was of why it was cheaper to ship some goods from China than it was to ship from state to state. I had recalled something about Trump taking some measure against this, but apparently at that time nothing had changed. I tried finding more info on Google, but I couldn't find what I was looking for. On DDG, I located an article from Newegg describing the UPU, and how these deals were established many decades ago and gave 3rd world countries at the time, such as China, a break. I then found another article from thehill.com that explained that Trump had begun the one-year notice of withdrawing from the UPU, but until then the rates would be the same.
On another occasion, I was in a thread that was talking about middle class wages, and I noticed that on Google, I could find many articles about how the middle class was shrinking. On DDG, I found some of those articles, but also found information on how if you control for inflation and hours worked, middle class wages actually increased since the 90s.
>"2020, certainly on top of my old organization, Trust and Safety, top of mind, they've been working on it since 2016, to make sure we're ready for 2020"
>"My definition of fairness and bias specifically talks about historically marginalized communities, and that's who I care about. Communities who are in power and have traditionally been in power are not who I'm solving fairness for."
>"Our definition of fairness is one of those things that we thought would be like, obvious, and everyone would agree to and it wasn't. There was, the same people who voted for the current president who do not agree with our definition of fairness."
Her definition of fairness is an ideological definition. It doesn't mean "fair," it means weighted, that is not objective, that is manipulated towards her ideological goals.
>"For example, imagine that a google image query for "CEOs" shows predominantly men... even if it were a factually accurate representation of the world, it would be algorithmic unfairness."
>"ML fairness... So they're trying to modify the model such that even if the data for female CEO says... its low... It still balances out."
There is really no doubt about the political ideology driving all of this, but what is apparent more than all of that is that they believe they have a duty to manipulate reality as much as possible and as needed, in order to fit their user's perception with their ideological narrative. And it is basically abusing their power (with ML) to enforce a delusion.
>"Alignment with the narrative."
>"People (like us) are programmed."
This is Ministry of Truth levels of social engineering/programming. They don't want to have to defend their ideology, they want to eliminate all opposition, and I don't know how any free thinking individual can defend that, much less support it. Its totalitarian, tyrannical, and the most terrifying aspect of it all is that they are knowingly committed to the construction of an entirely fictional, delusional narrative version of reality.