Machine Learning: Will it create a more or less equitable future for individuals? | C2C Community

Machine Learning: Will it create a more or less equitable future for individuals?

  • 10 June 2021
  • 7 replies
  • 76 views

Userlevel 7
Badge +26

Machine Learning can determine trends long before they are known to the general public. The U.S. real estate market is a great example of machine learning in action. On one hand, ML has created efficiencies that benefits many people. On the other hand, what are the long term implications of trend forecasting on individual consumer buying power?

Will machine learning create a more equitable future for all individuals?


7 replies

Userlevel 7
Badge +26

@tuliani @hifce I’m curious what you think about this.

Userlevel 3

I think ML can certainly help create a more equitable future but the usage of ML is in human hands so it’s not always straightforward. As far as Google goes, there are strong AI principles used to evaluate each use case. I know the White House had established some guiding AI guidelines as well; not sure if they are widely adopted yet.  

Userlevel 7
Badge +26

I think ML can certainly help create a more equitable future but the usage of ML is in human hands so it’s not always straightforward. As far as Google goes, there are strong AI principles used to evaluate each use case. I know the White House had established some guiding AI guidelines as well; not sure if they are widely adopted yet.  

@Andrada.Morar do you know of an example of an AI principle that specifically keeps individuals in mind?

To follow this conversation further, check out this this C2C talk about mapping AI ethics:

 

Userlevel 3
Badge +2

complex issues: we create systems to help people and relieved businesses. But AI scares both employees and managers (untrained). Ethics do not arise, because the people of the profession that I know (do not want to do harm, there are few of us in France who are really into machine learning).
The main concern is the data, its orientation and its protection. A tensorflow model, as ethical as it is if it is made to ingest data that is not, so it is not ethical.
As good as the wishes are, if we are bothered from the start, we will not be able to help our customers

Userlevel 7
Badge +26

Thank you for this insight @Darklinux!

Userlevel 7
Badge +65

I remember, almost two or three years ago, when I started playing with ML, reading an article about how life insurance went up. Because companies created models that could predict if someone would have health problems.

On the other hand, we have seen how ML-AI helped our health predictions.

So, we can clearly see, once again, that it’s up to us how we will use it. If it’s going to bring a more equitable future or not.

Reply