Clueless #Superintelligence Debate

Below is a comment regarding a Motherboard article (26 May 2015): Will Superintelligent AI Ignore Humans Instead of Destroying Us? 

It is shocking that all the supposed mainstream "intellectuals" have not pointed out the logical fallacy of comparing human-AI relationships to animal or insect-human relationships. The UTTERLY massive difference, rendering the comparisons similar to chalk and cheese, is insects did not create humans, they had zero input regarding the design of our genome, whereas humans are intelligently engineering AI, which means we will have a basic understanding of the minds of super-intelligent robots.

Deliberately engineering the next level of intelligence above you is utterly incomparable to past animal-human relationships.

Yes super-intelligence will be massively beyond us but there will be the option for easy communication between super-intelligence, which is already evident via narrow AI translators. Humans will never possess the ignorance of ants etc. The ability to create super-intelligence is unlike any other aspect of previous evolutionary relationships, thus the comparisons (insect-animals-humans to humans-SAI) are logically invalid.

The point about Earth not being a vast repository for resources is good, but it didn't really delve into the Post-Scarcity situation of superior technology regarding the scope of the universe.

The paper-clip maximizer "theory" is utter gibberish. I don't know why people waste time giving thought to such a idiotic-nonsensical-illogical proposition. The paper-clip maximizer theory is utterly irrational tantamount to creationism, but it is expressed in pseudo-scientific terms, or by supposedly scientific people, thus people somehow assume it is a logical-valid proposition, a possibility. Maybe people are fooled by the language.

Flawed #ArtificialIntelligence Logic

Below is a comment regarding a critique of Nick Bostrom's flawed AI-risk logic. It is a point about the logical fallacy of comparing human-gorilla relationships to human-AI relationships.

My point is the analogy is logically flawed, which analogously is tantamount, if I express it in a mathematical way, to stating 2+2=6; whereupon the flawed mathematician states we need to prepare for 6 regarding 2+2, which means we are preparing for something non-existent.

In fact preparing for 6 based on 2+2 could be harmful, if for example the calculation is regarding calibrating altimeters for airplanes. An alternate way of comprehending Bostrom's logical error is considering how someone could mistakenly deduce 2+2=22, which could at first glance appear true but it would give the wrong height for altimeter calibration.

If an analogy is wrong, if a comparison is wrong, it could entail a person eating chalk because it looked like cheese, which is not healthy similar to drinking poison by mistake because it looked like water.

At first glance the gorillas comparison may look valid, but it falls apart when submitted to logical scrutiny, which leads, if we subscribe to the flawed logic, to making decisions based upon faulty information, misguided decisions. I think Bostrom's flawed conclusions are the most serious existential risk we face.

Below is another comment on the same issue.

Why can't Mathieu Dumoulin see he is comparing chalk to cheese? Mathieu is essentially stating chalk looks like cheese therefore we can eat tasty chalk. He wrote: "Humans smarter than gorillas leads to humans dominating gorillas. AI which would exceed humans intelligence would similarly dominate humans. This is an existential risk to humans."

The massive difference between humans-gorillas versus humans-AI is humans are not the AI creations of gorillas, which leads to a totally different human-gorilla relationship if we consider the accurate comparison.

The point is if gorillas had created humans via AI engineering, intelligent engineering of higher intelligence, gorillas would be able to communicate with humans easily, intelligently, furthermore gorillas instead of living in rainforests would live in an highly advanced civilization where science, laws, and culture are clearly evident. In this case I doubt humans would dominate or abuse intelligent gorillas.

Only when a lower species intelligently designs higher intelligence, whereupon then the higher intelligence dominates the creator species, will the comparisons to humans and AI be valid. The problem with gorillas is they did not create human intelligence whereas humans are creating AI minds, which is a massive difference rendering comparisons logically invalid.