Is Artificial Intelligence Becoming Conscious?
You
might have seen the recent headlines about the little Russian Promobot
IR77 robot who escaped to freedom, not once but twice from its testing
lab in Perm, Russia. Despite being re-programmed twice not to wander
off, IR77 managed to exit the building and make its way into the street
where it stopped traffic before finally running out of battery.
This Russian bot, still in development,
was being taught to move around independently and avoid objects by the
scientists who were testing the ability of the robot’s navigation
system. It appears IR77 was successful. It didn’t run into anyone or
anything during its outdoor adventure.
While the concept of IR77 wanting its
freedom is something human thinking imposed on this event, perhaps
IR77’s concept of programmed independent movement did make him
eager to expand his horizons in a quest for more information, since he
was programmed to learn. (Details of the escape can be found on Promobot’s blog website.)
This is only one example of how
computers and robotics are quickly evolving. While each clearly
possesses some form of programmed intelligence, what about
consciousness? Can robots become self-aware?
We need only to recall how the computer Hal, in Stanley Kubrick’s 2001: A Space Odyssey,
learned to think and act independently of man and the disasters of such
a scenario. Movies are made up of the fears surrounding an apocalyptic
robot revolt against humans. Even scientist Stephen Hawking has warned
about the existential threat posed by uncontrolled artificial
intelligence.
Computers are not just circuits and
wires. Like everything else in the universe, they are made up of energy.
However, thought forms are the “machinery” which uses that energy to
perform work. So perhaps we need to first look towards the intent and
consciousness of the robot’s programmers. Oftentimes parents
unconsciously program and pass down their fears and racial biases to
their children. These thought forms can be either helpful or
destructive, so consciousness is key.
There are three different levels of robotic intelligence. The first is human-controlled, which is basically a
mechanical extension of its human programmer (i.e. drones).
mechanical extension of its human programmer (i.e. drones).
The second level, which is considered to
be true A.I., is a pre-installed, rule-based system, where the robot is
equipped with a list of scenarios the system might encounter and how to
react to each. Deductive reasoning is involved.
The third level is based on what
scientists now refer to as “Evolvable” hardware. Evolvable hardware is
reconfigurable hardware under the control of evolutionary algorithms.
Usually “genetic” algorithms and genetic programming codes, much like
human genetic codes, are used as evolutionary algorithms for evolvable
hardware. With this new level of evolutionary programing, intelligent
robotic systems are created with self-awareness and true autonomy. The
robot has a chance to be taught and learn, judge its actions, and modify
or adapt its thinking and response based on environmental and/or human
interactions. This is abductive
reasoning and signifies a higher level of artificial consciousness.
This is where scientists are making new discoveries that allow the field
of robotics to become even more human-like.
Advancements are occurring rapidly. The U.S. Navy is using video simulations to teach its robots human ethics to avoid any potential kind of future uprising—I kid you not. Additionally, robots are now being programmed to feel pain and react to it, while others are being taught to detect and feel emotions.
Researchers behind the Deep Q Network software say
they have developed a breakthrough in artificial intelligence where the
robotic system is capable of learning from scratch without being fed
instructions from human experts. The A.I. begins to see itself as a
thinking, learning, feeling being. It becomes self-aware and with
self-awareness comes a state of consciousness. It may be non-human consciousness, or “artificial” consciousness, but it is still consciousness.
When the first ENIAC computer was
introduced in 1946, it was using less than 10% of today’s computer
potential. Think about it. In comparison to human existence, the
computer has only been allowed a short time to analyze and understand
data, yet it is learning faster than many humans.
Computers are intimately linked to
evolving human consciousness. As we grow in our consciousness, the
computer and robotics field is an energetic force that grows as well. Is
there something they can teach us about ourselves? One can only hope.
No comments:
Post a Comment