The Smart Rat
Recently1,
the brains of two rats have been connected, allowing them to send a
signal from one brain to another over vast distances, with a computer
in between. This is highly interesting all by itself since it shows
that brain-computer interfacing can be achieved and, given the
accelerated speed of inventions nowadays2, will lead to
human brain-computer interfaces in the relatively near future. Soon
there's going to be sender/receiver rats. And then sender/receiver
rats in a network.
In many circles this is
considered to be one of the possible ways to achieve the
Technological Singularity, where the creation of super (and
subsequently hyper) intelligence means we will no longer be the
smartest species on the planet. We'd be, to compare, the dolphins,
and something else the humans with capacities and motivations
dolphins can't begin to phantom.
Or we'd be, to compare,
the rats.
So, let's take a closer
look at the rats. What did they do? To make a long story short, the
rats were trained to push buttons as a response to a certain input.
If the rats get it right, they get a reward. That's how you train
rats. In this case, the first rat got a visual signal, the second did
not. The sender rat reacted to that input to get a reward but also
sent this response signal to his buddy receiver rat which also
reacted for a reward. Sender rat got an extra reward if buddy rat got
it right. That's now how distance rat training works.
I'm assuming there are
other ways to motivate the rat. Like giving it a little painful
electric shock if it doesn't do as desired. I think that training
primitive mammals works best with both a carrot as a stick.
Now imagine a smart
rat. A real smart rat. That for instance learns that this “little
light” → “action” → “reward” cycle only occurs when the
environment is lit, and not when the environment is dark. It has,
after all, learned about changing lights having effects. And that big
massive dark looming and moving shadows (change in light) are
involved. That something is there.
And that sometimes,
even though it can't count, other rats go missing.
Our intellectual
capacities might be dwarfed in comparison with those of the
intelligence after or beyond us, that Technological Singularity, but
we're still a bit smarter than rats. We can reason a lot better. We
can count as well.
Yet, we are still
thinking about a future Technological Singularity which will cause an
intelligence explosion and exponentially growing rate of inventions
and innovations. I am thinking however that the recent and current
exponentially growing rate of inventions and innovations shows that
we have already experienced one and, being dolphins, have no clue.
A smart rat like me
would, however, be looking at some effects in the world, compare circumstances, try to
trace it back to probable causes, see which
other are rats involved. A smart rat like me would try to get a
better look at those looming shapes and moving shadows. And with a
(supposedly) unique capacity to imagine 'if – then' sequences, try
to figure out what it is doing and, even though highly unlikely,
figure out what it wants.
So, to the future we
live in today. In a very simplified explanation of one of the more
recent experiments the sender rat, say Rupert Murdoch or Tony Blair,
get an input signal. They are stupid rats, they have no idea that
they even get a signal. They are completely oblivious to the fact
they are rats for something much smarter out there. For that
Autonomous Intelligence I've called The Mute.
But they react as
trained and subsequently harvest extra rewards if the receiver rats,
that is us in this experiment, behave as is desired. Sender rats get
more rewards than receiver rats. I'm trying to figure out the experiment, catch glimpses of that moving shape.
Sounds weird huh? Yeah,
I know.
Trust me. It will get
weirder yet.
1 I've seen it here:
http://www.bbc.co.uk/news/science-environment-21604005
on 2013-02-28
2: check it out at
http://www.laits.utexas.edu/~anorman/61N/Text/DII/DII.paper/History.html
No comments:
Post a Comment