The Indian American

Amazon has yet to share the data used to train its system, signaling a major red flag, as most existing facial recognition systems suffer from a relatively high failure rate that results from racial and gender bias. Additionally, the challenges involved in cre- ating greater transparency immediately arise with the dark screen that divides con- sumers from the algorithm, which is further reinforced when we, out of convenience and haste, sit back and rely on the machine to make decisions without questioning it. A major problem baked in this situation lies in the huge gap in knowledge that prevents most of us from comprehending the algo- rithms that are being used; however, this does not mean that we should resign our- selves to accepting the unknown and trust- ing that machine will always make the best decision for us, as was the case in the Uber scenario. This knowledge gap is further exacerbated by the fact that corporations are unwilling to share their data-aggregating and process- ing methodologies in order to retain com- petitive edge in the market. If Google were to reveal the technical secrets to its highly successful algorithms, it would no longer reign as the primary search engine of choice. Eli Pariser, author of The Filter Bubble, has long railed against the intrusive privacy policies of Big Tech; he states that because corporations are so invested in monetizing the data that they collect, that “There’s not even a real basis to establish objective research about what’s happening on Facebook, because it’s closed.” Finally, as AI-driven products become woven into our daily lives, and as they gain more human-like qualities, as is the case for Siri and Alexa, we have to be more con- scious of how we interact with this technol- ogy. Julia Fink’s study in Infoscience states that “One approach to enhance people’s accept- ance of robots is the attempt to increase a robot’s familiarity by using anthropomor- phic (humanlike) design and ‘human social’ characteristics.” AI is most powerful when it is designed as something familiar to us. In such cases, we have to actively resist the temptation to accept the superficiality of its familiarity and recognize the logic fueling the algo- rithm. In Frankenstein’s case, the monster bore destruction on society because, although it was human-like, it was also far more power- ful than a human being, making it an inde- structible misfit among its surroundings. So, as we bask in this exciting era with the great potential for good facilitated by AI, we also must be less focused on the ends, and pay closer attention to how our new-found technologies will impact our society on a global, socioeconomic scale. Ideally, policies need to be placed early on to enforce cooperation between corpora- tions, governments, and citizens in a mutual effort for greater transparency to make AI an advocate of ethics in society. 24 THE INDIAN AMERICAN OCTOBER-DECEMBER 2018