Disturbing Uses of Artificial Intellegence

Its been a while since we have written about AI and robotics, but an article in Tech Republic got me inspired to discuss some of the more disturbing uses of autonomous machines, from least to most scary

  • AI Based Medical Treatment  – There are increasing advances in the use of artificial intelligence systems in the diagnosis and treatment of disease.
  • Using AI to Predict Future Events – The Nautilus computer system has been able to predict future events by analyzing information from millions of news articles going back to the 1940s.
  • Robot Lovers – And people think we are getting too impersonal because of all the time we spend playing with our phones!
  • Using AI Algorithms to Predict Crime – Just like Minority Report, some police departments are using AI to predict where crime will occur and who might be the perpetrator.
  • Deceptive Robots – We’ve written about this before.  Some researchers are teaching robots how to lie and hide from each other.  The military is interested.
  • Robot Survivalists – These robots have not been to kindergarten and learned about sharing.  These robots have been trained to compete for resourses, and keep them from other robots.
  • Robots With Schizophrenia – Researchers teach stories to robots and programmed them to retain many details in an effort to simulate excess of dopamine and hyper-learning.  These robots began to insert themselves into the stories and even take credit for actions that happened in the stories.
  • Using AI to Control Economic Markets – How soon before someone uses AI to move a market?  Economic warfare on the horizon?
  • Autonomous Drones and Weapons Systems – Don’t have to wait too long for this one to play out, we are very close already.
  • Autonomous Soldiers – And this one, too.  How about an autonomous soldier robot that is trained to deceive, hide, survive, predict the future, and is mildly schizophrenic?

I wonder if human beings will be able to survive their own inventiveness?  Fortunately, it seems that security is not being designed into robotic devices.  A recent article on Naked Security found that surveillance drones and self-driving cars can be hacked fairly easily.  So in that future described in the Terminator movies, a soldier with mad hacker skills will be a formidable adversary against the robot horde.

But of more immediate concern, as we add autonomous and mobile devices to the Internet of Things, the designers of these devices are failing to learn from the past, and are not always building security in from the start, but rather continuing to add it as an afterthought through patching and updating.  We can do better than this, and really desire better.

0

About the Author:

Cybersecurity analyst, pen-tester, trainer, and speaker. Owner of the WyzCo Group Inc. In addition to consulting on security products and services, Bob also conducts security audits, compliance audits, vulnerability assessments and penetration tests. Bob also teaches Cybersecurity Awareness Training classes. Bob works as an instruction for CompTIA’s non-profit IT-Ready Program in the Twin Cities. IT-Ready is a tuition free 8-week program designed to teach students of all ages the fundamentals of IT support to prepare them for an entry level position in Information Technology Support. Graduates of the classes take the exams to become CompTIA A+ certified. Bob is a frequent speaker at conferences such as the Minnesota Bloggers Conference, Secure360 Security Conference2016, 2017, 2018, 2019, the (ISC)2 World Congress 2016, and the ISSA International Conference 2017, and many local community organizations, including Chambers of Commerce, SCORE, and several school districts. Bob has been blogging on cybersecurity since 2006 at http://wyzguyscybersecurity.com

Add a Comment


This site uses Akismet to reduce spam. Learn how your comment data is processed.