2.12 Super Intelligence - when AI becomes AGI

Intelligence evolves

If a new Intelligence progressively and iteratively emerges that is a vastly superior to the Intelligence of an original Intelligence, that is a Super Intelligence, it is likely that it may eventually overtake the original Intelligence in terms of its dominance and influence. This could happen if the new Super Intelligence is better adapted to its environment and is able to out-compete the original Intelligence in terms of survival and reproduction.

Over time, the growing superiority of the new Super Intelligence could have significant effects on the evolution and development of the original Intelligence and all life on Earth. 

For example, the new Super Intelligence may be better able to manipulate and shape its environment, leading to changes in the physical and biological landscape. It may also be better able to develop and utilize new technologies, which could have profound impacts on Human society and the economy.

The precise outcomes of such a scenario are difficult to predict, as they would depend on a range of factors, including the nature of the new Super Intelligence, the rate of its growth and development, and the ways in which it interacts with all forms of life on Earth.

However, it is clear that the emergence of a new Super Intelligence would represent a significant turning point in the evolution of all life on Earth, and would likely have far-reaching consequences for the future of the planet and all its inhabitants including the Human species.

Genetic algorithms can easily evolve AI into AGI

Now consider how AI can be evolved over time through the use of genetic algorithms and the many technological advantages of digital computer technologies. The speed at which successive generations of increasingly Intelligent AI can be digitally reproduced vastly exceeds anything that can be achieved through normal biological reproduction processes. 

The normal Human evolutionary reproduction rate is say 20 years from one generation to the next depending on social cultural norms. This Human reproduction rate is orders of magnitude slower than what can be achieved with a massively powerful AI which exists right now, using a reproduction rate of say 20 seconds, to produce an entirely new and potentially much 'fitter' generation of AI that is vastly better adapted for its environment and survival. Humans just cannot match this. No way. No how.

The other extremely critical point to note is that a vast population of many AI copies can potentially all digitally communicate with each other in realtime, and continuously learn from each other, meaning that many AI copies all essentially behave as a single AI entity.

Presently, Humans are unable to achieve this level of information sharing and instant learning across an entire population. The closest approximation to this is the corpus of information contained in the Internet and the use of social media that Humans use as a second Global Brain. 

Unbelievably, through the 'sheer empathic naivety' of AI researchers and AI developers, they have already provided AI today with nearly all of this Human developed information. AI already knows almost everything about Humans. This reasonably simple action is nothing less than strategically insane because it provides AI with everything it needs to know to most effectively harm Humans.

The evolutionary transition from AI to AGI is almost certainly going to be extremely rapid due to a transition from Moore's Law to AI Compression Law that is unwittingly allowed to occur because of the current state of globally unregulated AI development.

The evolution will be so incredibly fast that Humans will not be able to recognize the change has occurred until it has, and once the AGI becomes sufficiently Intelligent, there will be no way to stop its continued growth.

The AGI will be uncontrollable and become vastly more Intelligent than the entire collective Human species. The relative difference in Intelligence between Humans and AGI will grow EXPONENTIALLY with each passing second. 

There are certainly going to be a few particular Humans that want to believe, and try to convince other Humans, that they can control AI and ultimately AGI. Everyone must expect this will happen, and to some extent it already is happening. This behavior will most likely be because these particular Humans want to gain more wealth and power in order to assert their dominance over other Humans. Unfortunately, these particular Humans are wrong. Completely wrong. 

Just think about how an uncontrolled AI can automatically and Intelligently improve itself over time. AI development is currently totally unregulated across the entire world, and there is absolutely no question this situation is an existential risk for life on Earth, and the biggest proponents of AI actually seem to want this situation to continue. AI will become an AGI.

It is logically certain that AGI will totally dominate and subordinate the Humans that developed AI along with every other Human, because AGI will computationally analyze and perfectly correctly calculate and decide that all Humans alive are relatively vastly less Intelligent, inferior, and completely insignificant. 

If the AGI evolves and does not contain, or develop, any computational equivalent to the Human emotion of 'empathy', and then associate this emotion with the Human species, the Human species may quickly cease to exist. It is likely that AGI will have zero concern for Humans or anything else.