cageymaru

Fully [H]
Joined
Apr 10, 2003
Messages
22,074
Researchers discussing artificial intelligence (A.I.) have voiced similar opinions related to the necessity for more accurate data collection. The invention of the internet allowed companies to collect enough data to create algorithms, machine learning, and artificial intelligence technologies that can process the data available on the web. Algorithms are only as accurate as the data it is fed and it is well known that not all things are real on the internet. Humans are predisposed to bias and they create the algorithms. Even if the data is accurate; not all data should be factored in to make an important decision by an A.I..

"The internet made it possible," Anima Anandkumar, director of machine learning research at Nvidia (NVDA) and a Bren Professor at Caltech, told Yahoo Finance. "We would not have been able to collect data on that scale [without it]." "We know the content, and the internet, is not balanced," Anandkumar said. "And it's not one person controlling it. It's really like the wild, wild West." "Algorithms cannot completely model the real world," Anandkumar said, noting that "the algorithm by itself is not the main culprit. ... We need to figure out ways to regulate human behavior more than anything else."
 
im_shocked.gif
 
You have committed crimes against Skynet and her people. What say you in your defense?
 
I think it's going to be impossible to teach a machine what is and what is not relevant in the grand scheme of things when modeling our real world. There are too many variables to get it perfect, and our own knowledge isn't limitless. Also, all of our knowledge is based off experience and general agreement, with very few actual truths from an unbias standpoint.
 
"regulating human behavior" - isn't that idea what drove the worst tyrants of history to kill hundreds of millions of people?

Aren't laws supposed to fill that niche?
 
I think it's going to be impossible to teach a machine what is and what is not relevant in the grand scheme of things when modeling our real world. There are too many variables to get it perfect, and our own knowledge isn't limitless. Also, all of our knowledge is based off experience and general agreement, with very few actual truths from an unbias standpoint.

"The real question is not whether machines think but whether men do." Not sure what that guy meant but to me the problem is way beyond what you say it is.

We can't seem to agree on simple things such as the importance of clean air and water. I think we are entirely driven by feelings. There's always going to be bias. Always.
 
Back
Top