Unbelievable! Bias In Artificial Intelligence Too

0
1027

Artificial Intelligence is used as an aid in every field.  Every sector healthcare, infrastructure, finance, automobile, hospitality, etc AI is in use. Nowadays most of the organization’s use AI it’s as if like a trend or something that’s so essential in running a business. The organizations using AI is increasing day by day as new technologies are coming up. Even the current situation prompts us to use AI, because the pandemic has created an alteration in the working style and most of the employees are working from home, various roles that require physical presence can be substituted by the use of AI.

             AI has a bad impact in human. There are proofs why AI is has a negative or bad implication towards humans. The incidents like Microsoft’s row over mislabelling a famous singer which was a mistake happened because of the biased data fed in AI. George Floyd’s death brings out that the technologies we use are biased, as the history of AI has evidence or proof of racial and ethnic bias.

              The bias of AI happens during the three stages – building the model, data collection, preparing dataset. The most common bias happens when the data is collected. The data gathered may not be based on the facts or a pre convinced decision. The systems can be trained accordingly based on the data collected and inbuilt. The facial analysis rate has high errors or bias due to the data inbuilt. For example – when amazon was hiring employees the systems eliminated female candidates it was because the system was trained in a historical perspective that male candidates should be preferred.

                The possible solution to remove bias is by changing the algorithms inbuilt in the system with new algorithms and training the system. AI helps use in identifying the way or the perspective in which we are biased and even we need to change that perspective.  Researchers have found innovative ways to remove discrepancies in the system. Organizations should encourage more ethnical means and reducing discrimination.

          The most important part to be noted or reflected is that AI is just a tool, the main change or the main improvement should happen within ourselves which will bring in many differences and benefits along. So, we are the ones that need a change more the biased AI.