The latest advances in artificial intelligence (AI) have raised several ethical issues. Maybe one of the most crucial is whether humanity will be able to control autonomous machines.

It is becoming more and more common to see robots in charge of housework or self-driving vehicles (including Amazon’s ), that are powered by AI. While this type of technology makes life easier, it might also complicate it.

An international group of researchers warned of the potential dangers of creating overly potent and standalone applications. Employing a set of theoretical calculations, the scientists researched how artificial intelligence may be kept in check. His judgment is that it could not be possible, according to the study released by the Journal of Artificial Intelligence Research portal.

Artificial Intelligence

“A super-intelligent machine that controls the world sounds like science fiction. However there are already machines which carry out certain crucial tasks independently with no programmers fully understanding the way they heard it […], a situation that could at some stage become uncontrollable and dangerous for humankind,” said Manuel Cebrian, co-author of this study, to the Max Planck Institute for Human Development.

The scientists experimented with just two strategies to control artificial intelligence. One was to isolate her from the net and other devices, limiting her contact with the external world. The problem is, that would greatly lessen its ability to perform the functions for which it was established.

Another way to design a”theoretical containment algorithm” to ensure an artificial intelligence”can’t harm people under any conditions.” However, an investigation of the current computing paradigm revealed that no such algorithm can be produced.

“When we decompose the issue into fundamental rules of computing, it turns out that an algorithm which instructed an AI to not ruin the world could inadvertently stop its own operations. If this occurred, we wouldn’t know whether the containment algorithm could continue to analyze the danger, or whether it would have ceased to include the harmful AI. In effect, this makes the containment algorithm unusable,” clarified Iyad Rahwan, yet another of the investigators.

Artificial Intelligence

Based on the above calculations, the issue is that no algorithm can determine whether an AI will harm!! The researchers also point out that humanity may not even know when superintelligent machines have come because determining if the gadget possesses intelligence superior to people is in exactly the same kingdom as the containment issue.


Most researchers agree that a superintelligent AI is not likely to display human emotions such as hate or love and that there is not any reason to expect AI to become blatantly benevolent or malevolent. Rather, when considering how AI could become a threat, experts believe two scenarios most likely:

  • The AI has been programmed to do something devastating: Autonomous weapons are artificial intelligence methods that are programmed to kill. At the hands of the incorrect individual, these weapons might easily cause mass casualties. Additionally, an AI arms race could inadvertently result in AI warfare which also leads to mass casualties. To avoid being thwarted by the enemy, these weapons would be designed to be quite difficult to simply”turn off,” so humans could plausibly drop control of such a situation. This threat is one that’s current even with lean AI but develops as levels of AI wisdom and autonomy increase.
  • The AI is programmed to do something beneficial, but it develops a harmful method for achieving its purpose: This can happen whenever we fail to fully align with the AI’s aims with ours, which can be strikingly hard. If you ask an obedient intelligent car to take you to the airport as quickly as possible, it may get you chased by helicopters and covered in vomit, doing not what you desired but actually what you asked for. In case a superintelligent system is tasked with a demanding geoengineering project, it might wreak havoc on our ecosystem as a negative effect, and see human efforts to prevent it as a threat to be fulfilled.

As these examples attest, the issue about complex AI is not malevolence but proficiency. A super-intelligent AI will be extremely proficient at accomplishing its objectives, and when those goals are not aligned with ours, we’ve got an issue. You are probably not a wicked ant-hater who measures on ants out of malice, but if you are responsible for a hydroelectric green energy job and there is an anthill from the area to be flooded, too awful for your ants. A vital goal of AI safety research is to never place humanity in the position of those rodents.

Artificial Intelligence

Artificial Intelligence


Not wasting time about the above-mentioned misconceptions lets us focus on authentic and interesting controversies where the experts disagree. What type of future do you want? Should we develop lethal autonomous weapons? What would you like to occur with project automation? What career advice do you give today’s children? Do you prefer new jobs replacing the previous ones or a jobless society where everyone enjoys a life of leisure and machine-produced wealth? Further down the road, would you like us to create superintelligent life and spread it via our cosmos? Will we control intelligent machines or will they control us? Will intelligent machines replace us, coexist with us, or merge with us? What will it mean to be human in the era of artificial intelligence? What do you want it to mean, and how can we create the future that way? Please join the conversation!

Artificial Intelligence myths

Artificial Intelligence myths