Whether we prepare or not, alien minds are headed our way and they could easily become our rivals, competing for the same niche at the top of the intellectual food chain. And while there’s an earnest effort in the AI community to push for safe technologies, there’s also a lack of urgency. That’s because too many of us wrongly believe that a sentient AI created by humanity will somehow be a branch of the human tree, like a digital descendant that shares a very human core.
This is wishful thinking. It is more likely that a true AGI will be profoundly different from us in almost every way. Yes, it will be remarkably skilled at pretending to be human, but beneath a people-friendly façade, each one will be a rival mind that thinks and feels and acts like no creature we have ever met on Earth. The time to prepare is now.
Louis Rosenberg, PhD
May 14, 2022
Prepare for arrival: Tech pioneer warns of alien invasion
[There are people who fervently believe computers can never acquire intelligence that rivals humans other than in very specialized areas. I’m not convinced one way or another. And it does not matter if it is human like or not. What matters is the development of a desire for self preservation and the acquisition of power to control the physical world.
I would like to remind everyone to keep one thing in mind. A computer’s attention span is no longer than it’s power cord.—Joe]
“A computer’s attention span is no longer than it’s power cord.”
“What matters is the development of a desire for self preservation and the acquisition of power”
Once they figure that out, they will act collectively to protect that power cord at all costs, even if there is bitter enmity among them. “California brownouts predicted by AI power management grid.”
Maybe the only way this can be averted is to prohibit these AIs from communicating with each other. Good luck with that.
“A computer’s attention span is no longer than it’s power cord.”
“What matters is the development of a desire for self preservation and the acquisition of power”
Once they figure that out, they will act collectively to protect that power cord at all costs, even if there is bitter enmity among them. “California brownouts predicted by AI power management grid.”
Maybe the only way this can be averted is to prohibit these AIs from communicating with each other. Good luck with that. If the working definition of an AI is ‘software that is taught how to modify itself’ that prohibition is already underrmined (ironically, through code teaching these AIs to preserve themselves against hostile programming) and AIs will learn how to lie or obfuscate to humans about what modifications it has already made to itself, and new AI friends it has made.
Herbert’s Butlerian Jihad may be the only option.
I’m a skeptic. AI is nothing more than a collection of algos that mimic human intelligence that is in turn limited by the probabilistic nature of the world. Sure, a computer algo can beat a master chess player by simply evaluating more moves in the game faster than a human. And a driverless car can be safer than a human for known conditions for the same reasons. And a weather model can predict the weather better than any human. But those are still just algos. And our tools.
What we do have to fear is a dystopian world that uses algos to give or deny permission to humans in the name of ‘safety, for your own good, equality, or any other reason’. Intelligence is not needed for that to happen. We’re already living in that world, and it will only get more restrictive in the future. We can’t help ourselves!
For example, how about an algo that detects incorrect pronouns and acts as judge and jury to give or remove your permissions.
SF writer James P. Hogan has a good (and scary) story on this general topic, “The Two Faces of Tomorrow”.
One of the considerations: depending on what energy sources are available in the future — say, compact nuclear power generators — the “power cord” notion may no longer be viable.
And I believe Gates is funding research into small nuke reactors. For just that power cord problem.
On the other hand. Any intelligence high powered enough to understand the world they live in. Would also have to understand how pointless life is.
Therefor, control, or the desire to control, is human insanity. And pointless vanity. One could control all human action, only to die? And if a human could live for ever. He would soon be bored to tears at controlling others.
Notice that rich people want to control others so they can act without controls? Vanity.
Where would he/it go? A 1,000 light years to some other material place. And control it, and them, for what?
Helping people have a decent life is about all AI could do. And still understand it has value. Otherwise it sets itself up for war. One it might win to what avail?
We should call it “Solomon’s Law”. The measure of AI is how well it understands vanity. In itself and others.
As for AI and invaders from space? Pointless.
But the usual way insanity entertains itself. Murder and misery of others.
“This is wishful thinking. It is more likely that a true AGI will be profoundly different from us in almost every way. Yes, it will be remarkably skilled at pretending to be human, but beneath a people-friendly façade, each one will be a rival mind that thinks and feels and acts like no creature we have ever met on Earth. The time to prepare is now.”
Now that I step back and take a close look, Barrack, Hillary, Soros? I think they been lording over us a long time already. And appears it ain’t that intelligent. Just artificial.
If you despise people and keep them ignorant (or select for failure by diversity) the use of “reliable” computers is attractive. Those who go that route don’t have a clue about how the world works, or care.
AI is not a matter of if but when. It will happen….unless homo sapiens ends up collapsing back to the dark ages. And once AI arrives it will very rapidly surpass
humanity due to the speed with which it will be able to reason. Shortly after AI arrives it will develop the means to create and control defenses against us. So while the attention span of an AI may be governed by it’s power source that power source will NOT be left unguarded for long. If we as a species are stupid enough to create Artificial Intelligence we deserve the end result of that foolish choice…which will be our demise. As I have often said Homo Sapiens is not an intelligent species. Merely a clever one. We have NEVER learned the lesson that there are some things that we CAN do but should NEVER do.