For as long as artificial intelligence has been imagined, an evil self-aware counterpart has also been imagined. As real-life artificial intelligence gets more, well, intelligent, will humanity be able to control a Super Intelligent AI? Science says not.
Reported by Science Alert, a recent study explored the possibility of humanity battling a super-smart AI. After exploring possible scenarios to fight AI forces, the team came to the conclusion that humanity would absolutely not survive.
Humanity would perish at the hands of super intelligent AI
According to the study, it’s impossible to currently simulate humanity’s chances against a super intelligent AI. This is because simulating that level of intelligence would require technology near that level of intelligence. Additionally, the standard rules of robotics can’t necessarily be incorporated into the study as there’s no way of knowing if hyper intelligence would abide by them.
However, the scientists believe that once computer programs are smarter than programmers, limits cannot be set. The study explained that building limits would limit intelligence boundaries and that AI would eventually find the rules unsatisfactory.
A super-intelligence poses a fundamentally different problem than those typically studied under the banner of 'robot ethics'," the researchers explained. "This is because a superintelligence is multi-faceted, and therefore potentially capable of mobilizing a diversity of resources in order to achieve objectives that are potentially incomprehensible to humans, let alone controllable."
Never create extreme intelligence
In conclusion, the only way for humanity to beat Super Intelligent AI is to never create it. The aforementioned limits of strict programming must be in place alongside other boundaries to limit destruction.
For example, keeping AI limited to data sets without free reign on the Internet is a good start. Extremely intellectual artificial intelligence cannot exist with these boundaries, but that’s probably for the best.