Google MUM Search Engine – Is it Dangerous?

by Martin Armstrong

QUESTION: What is your opinion of this new Google MUM AI search engine?

HU

ANSWER: It is for complex searches, and it is using a neural net. What you must understand is that a neural net is capable of finding things within a database. It is not capable of original analysis. IBM’s Watson was touted as having the ability to cure cancer. It failed. It could beat anyone in Jeopardy, and it could play chess. But a neural net will never actually provide true thinking in an original context.

We are primarily funded by readers. Please subscribe and donate to support us!

I was creating AI during the early 1980s. I created a program to mimic a politician and let my kids play with it. It would store your answers and keep a conversation going. For example, it might ask if you have a pet. You respond, “Yes, a dog.” “Oh, boy or girl?” “It’s a girl.” “What’s her name?” “Napoleon.” It would record that info, so the next time my daughter would go on the computer, it would ask, “How is Napoleon?” She came home from school one day and saw I had the computer apart. She began to cry, thinking I kill it.

I learned how to create an AI system that would seem alive. It would have a personality. Curse at it, and it would curse back. But it wasn’t alive. It could not think for itself.  To accomplish the next level, I had to teach it how to analyze on its own. That could never be accomplished through a neural net. It takes more definitive power than just that. Still, it will never become conscious. It will not wake up one day and decide it hates my guts for some reason and wants me dead.

Google’s MUM will advance searching. However, it will not suddenly become conscious out of nowhere and rule the world. We have really evil people for that.

 

 

Views:

Leave a Comment

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