Human monitoring is crucial because the humans who invented AI require oversight from other humans, and those humans will require oversight as well. Engineers created AI bias, and engineers will be the ones to design the remedies to eradicate it.
Following the Pegasus Spyware software revelation, people from all walks of life demanded that software makers better govern and supervise their development. The reactions were predictable.
For years, there has been a well-documented distrust of technology and what it can achieve. Majority of polled, corrupt government officials were the most frightening factor ahead of the COVID, while technology was not specifically listed in the list of anxieties, over 28% of participants cited computers replacing people in employment as a key concern.
However, the fact that over 10% of people are afraid of zombies is a good indicator of how fearful they are of these monsters. Even still, the notion that individuals are truly concerned about losing their employment to AI, even so, the fact that recently called for moratoriums on the sale and use of AI systems until adequate safeguards are put in place demonstrates the seriousness of AI concerns.
Of course, other AI-based technologies DLP, RPA, text analytics, are included in NLP.
AI may be a force for good, assisting societies in overcoming some of the great difficulties of our day. However, if AI technologies are employed without appropriate consideration for how they affect people’s lives, they can have negative, even catastrophic consequences.
So, how should businesses respond, or should they respond at all, given the constrained circumstances in which they deploy AI? Despite calls for stronger regulation of AI and its use, many organizations have shown an appetite for tighter external oversight of bot use among technology and financial services decision-makers.
Because the humans who created AI require monitoring from other humans, those humans will require oversight as well as human oversight. The truth is that no matter how automated AI is or what data sources it uses, all components of AI were created by humans.
Engineers created AI bias, and engineers will be the ones to design the remedies to eradicate it. The current issue is that algorithms have the potential to absorb and perpetuate racial, gender, ethnic, and other social disparities. Amazon developers revealed that an AI model used to screen job hopefuls favored men over women in one case. Over ten years, the algorithm was trained using a company’s engineering recruits database.
Because the majority of the developers in the training data were men, Men are preferred by AI, therefore references like women’s team captain or mentions of an all-female educational school in a resume were devalued.
Had Amazon not discovered the issue, the AI system could have been widely deployed, reinforcing existing gender prejudices. A lot of examples were revealed based on it.
Follow and connect with us on Facebook, LinkedIn & Twitter
