We live in the age of the algorithm - mathematical models are sorting our job applications, curating our online worlds, influencing our elections, and even deciding whether or not we should go to prison. But how much do we really know about them? Former Wall St quant, Cathy O'Neil, exposes the reality behind the AI, and explains how algorithms are just as prone to bias and discrimination as the humans who program them.
Speaker: Cathy O'Neil
Animator: Niceshit Studio
Intro and outro animation: Cabeza Patata
Producer: Abi Stephenson
Transcript
You are welcome to link to, download, save or distribute our audio/video files electronically. Find out more about our open access licence.
Join the discussion
Comments
Please login to post a comment or reply
Don't have an account? Click here to register.
"Algorithms Are Opinions Embedded in Code" is about the most astute, simple, true statement anybody has made on the subject to date. I reference Cathy a lot in my day-to-day role.
The closest we come to the value lens currently in data terms is the schema or sub-schema. These are classically created a priori but future developments may create them in flight and the assurance of the same needs to be gated and expert judgements applied in an appropriate time frame to produce fair outcomes..
“We have to inject ethics in the process of building algorithms” - I cannot agree more with that statement but how to make it happen? If, let say, HMRC is working on an algorithm that is going to analyse my voice while I’m calling the customer service than how this process can be influenced?
The truth is that this is likely to always (or at least for the foreseeable future) be a two part process. The algorithim will do exactly what Cathy suggests, but it then becomes a question of looking at the output through a values lens. The good news, I think, is that we have demonstrated our ability to do this with quants over the years. Even before we started to use the word ‘Algorithim’ as frequently as we do now, we have often used values lenses to assess ‘raw’ data and then make our business decisions. I would suggest that the challenge is not new, but the scale of the challenge is new. It is more important than ever that we help people to talk about values lenses in increasingly confident ways.