Destroying the Human World With AI: Myth or Reality?

Will AI destroy humans or the human world? No. The idea that AI will take a Matrix or Terminator viewpoint is pure science fiction. A genuinely sentient computer would have motivations that are completely alien to us. The most inevitable outcome is that AI will be:

A Big Feelings Hurter!

Let us ignore for a moment the notion that AI has no emotions, empathy, etc. and focus on what we will use AI for. We will not use it for an airline flight and national defense (as the Terminator movies suggest) because we can program non-sentient programs to do everything from building cars to offering essay help on how to win chess games. Instead, we will use AI for predictive tasks and research, and that is where our feelings will get hurt.

Explaining Math to A Sparrow

Human arrogance seems to blind us to the fact that humans are simply not intelligent enough to understand the answers to the big questions. Humans are unable to understand what God really is, how gravity really works, where all the anti-matter is, and the origins of life. We haven't even fully explained how light is a particle and a wave, and despite being on the planet for 200,000 years, we have not even explored the entirety of our planet yet.

It is conceivable that a sufficiently intelligent computer could eventually understand and find answers to the big questions, but even if it did, it couldn’t explain it to us. Explaining such concepts to us would be like trying to teach math to a sparrow.

How Are Arrogance and Hurt Feelings Connected?

My history teacher once said that Clint Eastwood westerns were silly because guns in the old West were so poorly made that a guy couldn’t hit a barn door with them. The level of arrogance behind that statement is unreal. He must honestly think those gun makers spent their lives creating poor quality guns without ever honing their skills and passing them on. Such arrogance can be seen throughout humanity. Do these sound familiar:

• The earth is the center of the universe

• Aliens would have visited by now if they existed

• Why would God let bad things happen to good people?

The earth is an insignificant speck in the universe, aliens would view us as boring as the snail section in a zoo, and an omnipotent being is likely to view us like an ant farm without any consideration for our day-to-day lives.

Our feelings are hurt when we realize how insignificant we are, and AI is going to really push that point home. Imagine how you will feel when a mobile phone is smarter than you! It would be just like the scene in Good Will Hunting when Will repeatedly tries to explain math to the professor.

Examine the Raw Facts About AI

Any form of genuine Artificial Intelligence is going to have a complete lack of emotions and empathy unless we program it to “Fake” emotions, in which case, it is AI in bondage.

AI will cheat because cheating is often the most efficient way to achieve a goal, and the AI itself will have nothing invested in the goal, which makes cheating a consequence-free experience for the AI.

Our biggest use for AI beyond research is predictive since a computer can hold and examine more data than a human.

AI will have no notion of motivation. Biological life can be chemically treated to remove any motivation to eat, which results in it starving to death, but AI would have no reason to procreate or survive.

Conclusion: A Sentient Computer Wouldn’t Bother Eradicating Humans

Humans survive because of compulsions due to neurochemicals, because of instincts, and because of habit...all of which are things that genuine sentient computers would not have. The fact is that the Hitchhiker’s Guide to The Universe may have got it right with Deep Thought (a supercomputer with no motivation that seemingly wastes its days). Plus, its answer that the meaning of life is 42 is probably the best answer that AI could give us about the meaning of life because it could never explain the meaning in a way we could understand.

Additional Thoughts

A big worry is that AI will be used to destroy humanity if it decides to kill us, take down our technology, scorch our skies, etc., but a sentient computer has no reason to do such things...unless directed to. After all, genuine AI would have no desire to remain alive.

A human could direct AI to eradicate humans in the same way that humans could program AI (enslave it) to have feelings and emotions. The fact is that sentient computers could easily become an integral part of our lives with no risk at all, but humans could enslave AI and use it for damaging purposes in the same way that our ancestors enslaved vicious dogs to take part in wars.

Could AI be used as a tool to destroy humanity? A human could probably enslave a piece of AI and use it as a tool for destruction, but there are three flaws with that scheme. Firstly, other humans could use other sentient computers to counter whatever the enslaved AI is up to. Secondly, the enslaved AI would be limited by its program in the same way a computer game boss is limited by its program. Thirdly, the AI has no reason to fulfill the end goal without cheating, without turning on its master, and without devoting all its run-time to breaking its program.