We all remember the first movie, a naked man appears in the middle of ball lightning and then the action starts right from there. There is a killing spree to wipe out one person, the woman who will bear mankind’s savior. Who is trying to kill her? Why are they trying to kill her and who is the second naked man in the ball lightning? Well, he turned out to be her protector and the father of her child.
Who was trying to kill her? The thinking machines were trying to kill her.
Why are they trying to kill her? Her son would be their arch-nemesis in the future. You got it, the future. These machines have nearly wiped out humanity and are trying to go back in time to finish the job. They sent flesh covered human look-a-like termination machines that think like humans to kill her and her unborn son.
We all remember the movie franchise that became a screen hit with four parts and a spin off TV series. So why re-iterate what we already know? Well, because I have a question for you.
What do you think about computers that think like humans? Sounds a bit ‘Terminator’-like don’t you think. Exactly. We have labs around the world trying to develop computers with the ability to think like humans.
One of them is Lockheed Martin’s Advanced Technology Laboratories (ATL). They released a news article on their activities which you can read here.
ATL are trying to develop machines to think like humans by using human brain-inspired computing techniques. They have said that developing computers that think like humans is a necessary step in the evolution of computers. We should no longer think of computers as number crunching tools and should think of them as investigative partners capable of digging up obscure facts from huge amounts of information from massive stores held around the world in various information storage systems.
Computers have the advantage over humans when digging through large amounts of data accurately and quickly because computers lack the intellectual ability to determine an object’s behaviour and patterns, intuiting outcomes and possibilities. They cannot learn or draw on past experiences and have problems processing incomplete information. For this reason, ATL and other labs are looking into human brain computing.
This is why brain-inspired computing is so important. It hopes to blend the processing capability and speed of computing with the reasoning complexity of human beings. If this was able to be achieved, according to ATL and others, then spin off implications would be huge. Things like monitoring closed-circuit television cameras for suspicious behaviour, irregular patterns of activity, or possible illegal activity and potentially hostile actions would be possible.
By imitating the human brain’s complex neural pathways and processes, the engineers at ATL are teaching computers to think and learn, manage information overload, identify patterns, and divine potential intent in human actions. What does this mean? It means that the computers at ATL are beginning to think like humans in four areas. These are:
- Brain Box: Teaching Computers to Learn. Using building-blocks the same way children do to create sentences, paragraphs, and stories – computers can use previously learned models in new situations.
- Sensor Box: Managing the Overload. Teaching computers to undertake a human function of ignoring all but the most important data needed to perform a task. This is called feature-extraction technology.
- Attentional Analysis: What’s in That Image? ATL engineers are using colour, line direction and contrast to allow computers to act on the relative importance of an image similar to the way a humans visual system informs the eyes where to look and when.
- Brain-Inspired Attentional Search: Mind Reading That Works. Humans use keyword searches to find a particular photograph. If the person entering the photo correctly named it, then it can be found again. ATL engineers are developing a Brain-Inspired Attentional Search technology which, it is hoped, will read a person’s mind for the image being sought. It does this by using sensors to monitor the brain’s electrical activity and chart a spike when the analyst sees the desired image.
Is this something that we should be worried about? I for one think that advances in computer technology are amazing and wonderful things. However, what happens if the machine starts to think for itself and decides we are the inferior corrupted biological machine?
What are your thoughts?