To all participants: What change framework can you put in so that you can make an AI transition seamless and successful?

by bc3ef | Apr 30, 2021

Its important to make changes which are bite sized , explainable and hence digestible. Anything that doesn’t have an explanation or a reason will create resistance. SO removing the fear of unknown, explainable and with clear vision on what it helps with is the first...

Can AI be unbiased? The algorithms need to be created by humans and this will always include a form of bias.

by bc3ef | Apr 30, 2021

Yes absolutely. AI can definitely create bias. Machines will do what humans tell them to do. So if someone wants to create a biased or a sinister model they can. This is where I am sure just like we have GDPR or other certifications, we will have certifications and...

Question to the speaker, Nina: Any thoughts on the significance of masking names/blind contracts of appointment (similar to blind CV?)

by bc3ef | Apr 30, 2021

Absolutely, that is one of the great ways of avoiding biases. Answered by : Nina Alag Suri Founder and CEO, X0PA AI

Is the AI algorithm a one size fits all or do we need to build a different one for different cultures?

by bc3ef | Apr 30, 2021

This is again something where a lot of care needs to be taken to ensure, cultures do fall into personal data and should be avoided to ensure the risk of stereotyping and biases. However for eg in X0PA we take into account industry. Sector and regional data to make it...

How do you monitor if there is no bias creeping into your AI algorithm? If your historical data are for instance majority coming from a certain school or company will these historical data not create a biased algorithm?

by bc3ef | Apr 30, 2021

That’s a great question. Algorithms need to in first place ensure do not take personal sensitive information in its modeling. So for eg information around gender, age, ethnicity etc should be carefully removed from all models which is the case in X0PA’s models....