It's easy to take the title literally and object to it by saying "algorithms are neutral, they're just code". But Professor Noble isn't being that simplistic - her point is that in the real world, when actual humans are affected, algorithms are only as useful as the systems that feed them the data. It's not enough to create a "neutral" search algorithm and expect that the results it returns to you aren't going to be problematic for actual human beings, as shown by the real world example of typing in "black girls" and returning mostly porn. "Neutral" isn't necessarily neutral at all when it comes to real live humans, which is why Google and other tech companies are constantly tweaking their supposedly "neutral" algorithms in order to avoid garbage input data (i.e. society's harmful stereotypes about various groups) self-reinforcing the "truth" that's implied by a normal Google search. Remember the case where Microsoft's chat bot Tay got turned into a racist within hours? I'm sure Microsoft designed that system to be "neutral" too, but when it met the real world, things didn't go quite as expected. An algorithm doesn't exist in a vacuum, to be judged solely on its mathematical properties and nothing else - it's intended to be used, by *people*, for actual real human purposes, and therefore should be designed and judged in that context. If an algorithm, even unintentionally, leads to reinforcing stereotypes or creating harm to real human beings, then that algorithm is problematic and must be changed! And tech companies can do their part by actually altering algorithms in anticipation of how they are going to be exposed in the real world.
This seems a little much. What shows up is the most common searches. It means most people who are searching “black girls” are looking for porn. The searches people run are what teaches the AI. Saying “systemic” is new idea. She seems just trying to make a buck riding a wave of “racism”
I’m reading your book (first chapter so far) Yo mention several times how the book may become or the examples giving in the book may be outdated, but I believe it would be very much an important read for people who want to understand the true ramifications of the internet and AI. 5/5 🌟
This is beyond my expectations. I read a book with a similar theme, and it exceeded all my expectations. "Game Theory and the Pursuit of Algorithmic Fairness" by Jack Frostwell
I am in need of an accurately captioned version of this great video so we can include it in a course about digital citizenship for college faculty. Could you kindly enable the Community Contributions feature so I can edit the auto-captions? If you could please reply and let me know, that would be great. Thank you.
Shows exactly why students, including myself, are being brainwashed by this PC culture. I always stood up against people who chose other people's feeling over simple facts and data in college and school. If someone is offended by a simple figure, data or statistic it's upon them for being that close minded. Facts and reality can hurt, just accept it instead of changing it.
If you had basic critical thinking skills, you'd understand that's not what she's saying. These algorithms are the products of the mind who make them.Bias and ignorance undoubtedly is woven throughout.
Do search engines simply reflect...or do they direct what the public is interested in? These are the types of questions that E. Musk will answer by digging into Twitter algorithms, offering transparency, as he plans on doing.
It's easy to take the title literally and object to it by saying "algorithms are neutral, they're just code". But Professor Noble isn't being that simplistic - her point is that in the real world, when actual humans are affected, algorithms are only as useful as the systems that feed them the data. It's not enough to create a "neutral" search algorithm and expect that the results it returns to you aren't going to be problematic for actual human beings, as shown by the real world example of typing in "black girls" and returning mostly porn. "Neutral" isn't necessarily neutral at all when it comes to real live humans, which is why Google and other tech companies are constantly tweaking their supposedly "neutral" algorithms in order to avoid garbage input data (i.e. society's harmful stereotypes about various groups) self-reinforcing the "truth" that's implied by a normal Google search. Remember the case where Microsoft's chat bot Tay got turned into a racist within hours? I'm sure Microsoft designed that system to be "neutral" too, but when it met the real world, things didn't go quite as expected. An algorithm doesn't exist in a vacuum, to be judged solely on its mathematical properties and nothing else - it's intended to be used, by *people*, for actual real human purposes, and therefore should be designed and judged in that context. If an algorithm, even unintentionally, leads to reinforcing stereotypes or creating harm to real human beings, then that algorithm is problematic and must be changed! And tech companies can do their part by actually altering algorithms in anticipation of how they are going to be exposed in the real world.
This seems a little much. What shows up is the most common searches. It means most people who are searching “black girls” are looking for porn. The searches people run are what teaches the AI. Saying “systemic” is new idea. She seems just trying to make a buck riding a wave of “racism”
Algorithms are never really "neutral". A bias of some sort will almost always exist.
I’m reading your book (first chapter so far)
Yo mention several times how the book may become or the examples giving in the book may be outdated, but I believe it would be very much an important read for people who want to understand the true ramifications of the internet and AI.
5/5 🌟
This is beyond my expectations. I read a book with a similar theme, and it exceeded all my expectations. "Game Theory and the Pursuit of Algorithmic Fairness" by Jack Frostwell
Obrigada! Comprei o livro e acho que tenho muito a aprender com o seus escritos!
I am in need of an accurately captioned version of this great video so we can include it in a course about digital citizenship for college faculty. Could you kindly enable the Community Contributions feature so I can edit the auto-captions? If you could please reply and let me know, that would be great. Thank you.
You have to be kidding? This was a load of fake stupid crap
Shows exactly why students, including myself, are being brainwashed by this PC culture. I always stood up against people who chose other people's feeling over simple facts and data in college and school. If someone is offended by a simple figure, data or statistic it's upon them for being that close minded. Facts and reality can hurt, just accept it instead of changing it.
Visiting this video after the Twitter black stock image vs white person in the stock image fiasco.
10 years ago there was a problem. Do you think something has changed since 2009?
definitely. searches for 'black girls' today won't automatically bring up porn anymore. it's definitely changed since then
Racism has always been profitable. Power to define others and benefits of bias information $$$....
Now is the time to change solve this problem
can anyone recommend other search engines? :)
If this is what makes you concerned about Google, you have a very myopic view of oppression.
nice work
Let's write a book about computers and call them racist :/
If you had basic critical thinking skills, you'd understand that's not what she's saying. These algorithms are the products of the mind who make them.Bias and ignorance undoubtedly is woven throughout.
Do search engines simply reflect...or do they direct what the public is interested in? These are the types of questions that E. Musk will answer by digging into Twitter algorithms, offering transparency, as he plans on doing.
He doesn't plan to do that at all and if you believe that, you're either extremely naive or a sycophant.
Do you like being puppeteered by Meghan Markle lol
Sickening...
ok, boomer.