The Coded Gaze: Bias in Artificial Intelligence | Equality Summit
HTML-код
- Опубликовано: 21 авг 2024
- Joy Buolamwini, founder of the Algorithmic Justice League, presents at the Bloomberg Equality Summit in New York. #BloombergEquality
Stay Connected
Twitter: / bloomberglive
Instagram: / bloomberglive
LinkedIn: / bloomberglive
More From Bloomberg Live: www.bloombergl...
Her contribution to the conversation is immense.Great job
Ditto. Go sis! Loved your earlier TED Talk, too. Keep shining your light!
Terrific lecture on algorithmic injustice. Still timely and urgent. Thank you Joy Buolamwini!
What is algorithmic injustice? the problem she is calling bias is rather due to machine learning limited training data sets.
CHEERS to YOU JOY!
THANK YOU FOR TEACHING THE TRUTH
The lack of comments is because this knowledge is not shocking to black people especially to those of us who work in IT and already know that the coders are predominantly white. Thank goodness for people like Joy Buolamwini who can support her statement with empirical data.
That's because most black people aren't interested in the STEM field and many of those who try and study in school are accused of "acting white". You don't see that many Jewish basketball players either but you wouldn't call the NBC anti-semitic would you?
White males!!!
@@Akihito007 This is a stupid generalized statement. Most Nigerians.. ARE in STEM fields. When you say blacks you have to specify.
@@iniu1255 It's not stupid. The VAST majority of blacks in America and the rest of the world are NOT going into the STEM field. You have many more Asians, Indians and white Westerners who are.
ummm shouldn't you be more concerned that she is working to include you in the system now? I don't want a system like this AT ALL
YOU ARE AWESOME. A True inspiration. We need more of you! Thank you for bringing more clarity to humanity! 👍👀💯
Absolutely amazing, thank you so much for this incredible presentation.
This woman is amazing!
I really can’t believe this video has less than a hundred comments. WAKE-UP People
Very interesting, especially when intersectionality is involved, a great talk and very engaging.
Great talk about algorithmic injustice and facial recognition.
Loved this talk! Amazing!
I’m very surprised there’s no single comment after more than a year of this video’s posting. It’s surreal I’m a first. Wow!
Yep. LOL.
Great presentation.
👏👏 That was needed
I am sooo AWAKE right now, you can't dooze for a second. Stay on it Shalini and all you IT watchdogs!!!👊🏽
Bless you Joy, for your efforts. Your Brilliant.
Just a light problem maybe? What if the camera couldn't get picture because the face is too dark then? Did you test the system with a better lighting?
She's dope 🔥🔥🔥
this realy got me thinking. thank you
My women of the year for the next 10 years!
what is the tracking system she used in her documentary?
no offense, but it really looks like an After Effects work to be honest, so I would love to investigate the actual software and see it myself.
This is about light reflection not bias.
Four downvotes by racist bots.
I m black and do not support his characterization of limited training dataset as bias. the same system she is talking about if trained with black people will mislabel white faces or struggle with white faces
WOW.
Gonna talk about that all in a Courthouse 😉👍 ...whatever if letters are making a word....
like T E D . ... Or ...
C O R O N A , ITS MILITARY TORTURING.
Too bad she didn't use her talent to make change, instead of just talking about how things are "unfair".
The technology today was made collectively by millions of engineers, spending hundreds of hour. She'll have to be reborn 1000 times and make changes, still that won't make a dent. Same with AI and data collection.
@@devshrestha4088 if you don't mind, I would love for you to elaborate. Honestly.
@@BirdDawg1 Data is created by millions of people, and data from internet generally has a lot of bias. Technologies that move data were created by thousand of engineers working collectively like kafka, kinesis, and spark. She can best create a voice that may spark some change in mindset on those who control how data is collected. Or may be make some changes on the tools that collect and feed the data to AI models that learn from these bias (its called sigmoid function). Imagine if you teach a child for early age, every dark skinned person is a criminal, that child or AI model will always believe that in its mind.
@@devshrestha4088 I don't understand your logic, why doesn't she instead of complaining work on making a deep learning facial recognition software with datasets that focus more with people of color. If she is a coder who works with ml, which is what she is coming off as, she could probably do this herself. Instead of bitching about it, and wanting to get other companies to do it.
@@devshrestha4088 And if your logic is oh well the facial recognition software was made by 100s of developers and engineers, that shit goes right out the window when you see some of the shit people who are self taught and code for fun make. People can make neural networks and ai in the comfort of their own home nowadays, I assume she has money she could easily hire a team to help her and use any existing libraries to try and do what she wants to do. Things like tensorflow make doing something like working on facial recognition pretty simple even for people who didn't go to MIT to learn advanced mathematics.
Nothing to do with skin colour and everything to do with the fact that she and many many black women look like black men. Better fix it, the truth isn't real. Right?
o yeah u say she is a male hmmmmmmmmmmm
Is she actually a coder or just a "coder"? I see plenty of "coders" who basically don't even code at all they sort of just understand the basics of computer science and networking, but can't actually code. They are more like the people who analyze analytics and work with implementation of math.
what are you rambling about. a code is someone who codes. my problem with her is that she seems not to know that the problem she is facing has to do with datasets not some inherent algorithmic bias
Are you missing the forest for the trees? If I would building a house and you told me the floorboards were missing, would I be right in criticizing or disregarding your input because you don't know to operate a pneumatic airgun?
@@austinjohn8713 yes the ai, is being fedbiased data