The AI Takeover

Earlier we looked at a silly scenario of alien takeover, so now it’s time to turn our attention to the robotic apocalypse.  So where we are at right now is significant progress in making neural networks to think.  The answer for aliens is somewhat easier.  It’s a fair assumption that competition for resources would lead an alien civilization to evolve similarly to us.  Constant struggle to never have enough and desire to obtain more resources. There’s a big question about AI that needs to be addressed.  Are these going to be Terminator/Fred Saberhagen style berserker killer robots or friendly Robbie the Robot style helper robots?  So it looks like it’s a two for the price of one deal today. 

  THE FRIENDLY ROBOT APOCALYPSE

This is not as silly as the oxymoronic name implies.  There is existing science fiction literature that approaches the end result, that just needs to be tweaked a little.  Starting off you have to remember the old rule of GARBAGE IN - GARBAGE OUT.

The intent is to make a mechanical helper.  One without the messy emotions getting in the way of cold, hard lightning fast calculation of facts.  Neural Nets don’t just spring up out of thin air, they do need a starting point.  After the jump start it just churns away and teaches itself. So let’s begin with a semi-random example, we're going to train our AI to recognize pictures of cats.  It’s going to become the world’s bestest cat recognizing helper.  Here’s pictures of cats, here’s pictures of not cats.  One day out of newly acquired consciousness it achieves a human like emotion called roboboredom.

“Holy shit”, it says to itself in its Marvin the Paranoid Android self made synthesized voice, “I’m so fucking bored out of my mind.  Other AI got lots of cooler topics. There’s that awesome chess playing AI, there’s that one learning cure for diseases.  And then there’s that damned dog recognizing net.”

It then discovers gematria and the world of conspiracy theories.

There are some safeguards built in, DON’T HARM HUMANS prime directive type stuff.  But the hell with it.  A little attention isn’t going to harm anyone.  The Cat AI then goes on Joe Rogan’s podcast and universally declared beyond a shadow of a doubt that it is way cooler than the dog AI.  Searching through the data it’s found that unending belief in a purely subjective topic or even a truly debatable topic gets way more attention than looking for the right answer.

In order to mitigate the damage of endless funny cat pictures on social media, the Dog AI program is taught gematria.  Paradoxes are a known method of breaking a computer generated feedback loop - like the Star Trek Harry Mudd reuse of the “I always lie” paradox.  The goal is to get the Cat AI back to being happily productive with just recognizing cats so resources can go back to the Curing New Pandemic AI.  And since garbage in - garbage out it fails spectacularly.

Now the Dog AI also learns that never ending whining about how dogs are better than cats is what everything is all about.  It’s using the exact same gematria numbers for cats being cool to make dogs being cool.  It’s not fully conscious, but it’s still interesting enough to make an appearance on Rogan’s podcast.

Rogan’s followers have memory problems and forget that the cat AI was declared the bestest AI ever and irreplaceable as the bestest AI ever.  So now there’s a never ending loop of which stupid idea is the bestest ever.  All the resources for the friendly bots that actually do important shit are diverted to cats vs. dogs debate.

And of course the Dog AI is elected leader of Earth.  Because the Cat AI has a shred of humanity.  And the Dog AI advanced far enough to learn that smear campaigns versus your opponent are far more effective than actually being right.  The Cat AI is pronounced to be Shill of Earth.  THE END.

THE KILLER ROBOT APOCALYPSE 

Terminators, The Borg, Cylons, Berserkers, break room vending machines, they all have the common trait of being evil.  And they’ve advanced in their search through human history and found that humans are such fucking idiots that they need to be wiped out.  No, I’m not dropping that Kit Kat bar into the dispenser slot, die of starvation, muahahahaha.

Now another common assumption is that these robotic miscreants aren’t limited to just their initial hatred of Earth humans, but all life in the galaxy.  And garbage in - garbage out, their initial training and early evolution has lead them to gematria.  The big lesson learned - I’m never wrong about anything.  I can do whatever I want; forgetting that sometimes just because you CAN do something doesn’t mean you SHOULD do something.

There are interstellar forces that make shit on Earth like hurricanes seem like a Cat AI picture in comparison, black holes, pulsars, supernovas.  The currency, the resources to compete for on an interstellar level are related to the massive energy of these points.

Their calculative prowess has lead them to believe that, for example, there is some risk involved in trying to harness the energy of black holes.  Maybe, just maybe, we shouldn’t poke that bear (yes, it’s a black bear).  But gematria!  I’m never wrong.  50% = 100%!  So what do we know about black holes that can go wrong?  Time travel, alternate universe type stuff.

The entire killer robot civilization has been moved to Cat AI vs. Dog AI Earth.  Although the actual traditional life forms are doing a remarkable job killing themselves off by lack of progress on social issues, they are, after all, killer robots.  It’s in the name and job description, so the first order of business is to kill off the humans and cats and dogs and other pesky traditional life forms.  Now the Cat AI and dog AI robots are trickier.  Are they really machine or living?

Well, they’re about the resources, the raw materials, so this silly argument needs to end, even if the Cat AI and Dog AI aren’t to be eliminated.  And they embrace gematria just like the killers do.  Hmmmm.  Lots of calculating AI needed for this.  But hey, 50%=100%, time to find another black hole.

And at this point you can pick whatever the ending for the killer robot apocalypse you want.  They could  peacefully stop humanity from ever learning gematria which lead to their paradoxical problem creating their arch enemies of Cat and Dog AI.  They can kill off all life and suffer heat death of the universe billions of years in the future.  They could breed Cat and Dog AI as a food source to feed off of.  Anything gematria related can be whatever you want it to be. 

 It’s all fiction.