@wernerjacobs3282
Good job, explaining all these concepts really clearly
@THENOMADICAIKING
Great information and why everyone needs to stay up to date on what causes these problems, ones many do not realize are there.
@Sofiia-Fradynska_IT-Lawyer
Thank you for interesting videos re AI & ML.
@takeda0123
More I get to use and kind of understand current state of AI, more I can't help but think it mimics human development/traits very closely. It adds flavors to different AI. That keeps it interesting. But at the same time, as we will be getting into pitting my AI against your AI, it won't be any different from nations against nations, people against people, and civilizations against civilizations. Let the winner takes all.
@gonzaortin739
Truth is, you cant get rid of sample bias. Ai can improve, that is true, but I dont see current technology going as far as they say we’ll get. I know they must have hidden tech, but I see pretty hard problems to be solved to get to AGI, and I wonder if we will ever get there, when we still cant solve simple questions like identifying a rule for prime numbers. Maybe there is no rule for prime numbers, and maybe there is no possible AGI. We still dont know. The only thing I know is people usually overtalk to get attention and profit from that.
@CrispinCourtenay
Some of this bias is warranted, eg. driving into a sketchy part of town. Without a financial incentive, there would be little interest in doing so, safer areas that paid the same would be the first choice. Zip code, profession, employer make a difference, either from a human ranking or human trained AI. I would strongly recommend against over correcting here.
@thesimplicitylifestyle
“I’m not bad, I was just interpreted that way.” - Bad Data 😁 😎🤖
@debojitmandal8670
You forgot to mention one thing there is something called prompt engineering and there you can make the AI biased by giving it some special instructions for example is Arunachal Pradesh part of India and as per AI it's clearly biased of not giving a correct answer not to be hostile to other Nation there are AI bots that just clearly says that they don't have the answer for that even though the data was trained on English data and it's abundantly available and the data actually contains this information. So it's not completely true that the AI is biased because of the reasons you have mentioned below.
@sherpya
can't stripping demographic terms from data improve bias problems?
@kdjsgkhjhmrzazhgu
chainlink AI+Blockchain to solve this
@Newflame4422
Dr Joy Buolamwini
@gamesstreams9165
Just base words
@historia_tego_swetra
so, women don't use 'executed'? :D come on, that's idiotic
@NK-ne9uf
Cry harder lmao
@coreyleath4662
So does K means and Gradient Boosting regression have Algorithmic Biases?
@KavyanshPandey-l7d