The darker side of machine learning

By ITS Education Asia

photo credit: YO! What Happened To Peace? via Flickr

I came across this article recently while I was doing some other research and thought I would share it here on our blog and discuss some of the issues raised.

The darker side of machine learning

The overall article is discussing some of the recent applications of machine learning and puts a bit of a ‘sinister’ twist on how the technology is being used. However, the peaceful and useful applications of this technology could bring many benefits to us all.

The first app discussed is a new facial recognition app developed in Russia. This app can identify people based on using the database of images from specific social network sites. The positive applications of these types of apps are also likely to see many changes. Imagine a doorway greeting you by name as you enter a shop. Tailored response to your needs can be arrived at before you even ask at the counter. Security codes and passwords could become unnecessary as you will be identified straight away by the camera. Your information stored on the cloud will be instantly available to you from any data terminal. Passcodes at ATMs will no longer be required.

This could make being a criminal extremely difficult as the person with a particular record could be identified as soon as they enter a business. Stores could have, for example, databases of known shoplifters. Whereas those with a good spending history could receive VIP treatment as soon as they enter the business.

The overall effect is it could make dishonesty much harder and targeted marketing much easier. This could see productivity improve and aid economic growth. The need to have multiple different and possibly fake IDs would be eliminated. Those on terrorist watch lists would not be able to get on planes.

The second app discussed is the algorithm that imitates handwriting. This in itself helps explain the need for better ways of identifying people other than signatures. This type of app is going to make written signatures much less secure and will speed up the need for better ways of confirming identity – for example facial recognition apps.

The final app is my favorite. ‘Chatbots’ that can impersonate real people and engage in live conversations with you that mimic what the actual person would say. What a ‘hoot’ this could be. Imagine being able to call up Sheldon Cooper from Big Bang Theory for a discussion. Talk with an ex-president or your favorite music idol. These chatbots are likely to be a huge success. The ability to have a discussion with a departed loved one is another possible application that could help people with the grieving process.

To me I embrace the technological advances that we are seeing in this article. We cannot go back to the way we were before technology therefore we must move forward. These types of machine learning apps could bring great benefits to us all.