It was 2 am, and as usual I was half attentively scrolling on Instagram when I clicked on a post about not sleeping at night and just wasting time on social media and an ad popped up from IKEA that said it’s late you should sleep, is your mattress not comfortable enough? Come to IKEA we have a solution for your sleepless nights. Now it’s good marketing on your part IKEA but now I have anxiety! Can you imagine 2 am at night and this happened!
Well algorithmic predictions are ubiquitous these days, you could have googled about an essay for your homework and later on seen ads recommending related books on your social media. Nowadays all major companies use machine learning algorithms and bots, which can be very useful, but it can also reflect human bias like Amazon’s AI did while rating applicants, it scrapped applications from female candidates. Kay Mathiesen, an associate professor of philosophy and religion at Northeastern University says, “data is biased – because it’s data coming from human beings”. So now the question is can these biases be corrected? Mathiesen says that, there is potential to correct these biases. She says machine learning has the ability to both consider the most relevant factors in a decision and produce more tailored analysis. It can be much more objective than people. It will outdo the ability people currently have to process situations and people on their own.
So yes it may be possible that someday machines will be unbiased as they are supposed to be, well we are definitely working towards it so there’s still hope. For time being we will still be creeped out by seeing ads related to things we were just talking and never even typed on our phones. Like please internet stop terrifying us!
Mayuri Talgaonkar