How Dangerous Can AI Be?

Artificial Intelligence, AI, is widely used nowadays in the modern era. Although it does not look as sophisticated as the talking hologram you usually see on Hollywood movies, it has more or less similar function to them. AI’s job is to solve its own problem. It means that AI can learn from experience. Now doesn’t that mean they can also be a thread like, once again, that in Hollywood movies? The answer is yes. But it is not a matter of yes no question, but how?

Autonomous Killing Machine

codeWorld War II was indeed ended, but that does not stop or hinder the development of weapons. Many countries learn from the past and have started to build their own weapon, either defense or attack, for the incoming war. One of the weapons is a battlefield robot.

Battlefield robot is designed to get in the battlefield and take action whether to kill or not kill somebody. The problem is, the robot will make its own decision without human interference. The decision making is based on certain language coding that is put in the robot programming, but what if it flawed and the robot instead shoot a friend, not a foe, or kill a civilian, not an army?

Data Mining Tools

AI2It was some years ago that Americans’ took an action of demonstration because their personal information was mined by the government. They were in rage because the government violates their privacy, break into their emails, phone records, and other personal data and information.

It was accused that such action was carried out with the help of an AI. All the data, which was so huge for human brain to analyze, was sorted and organized in a certain way only an AI can do. Now that can happen again in the near future, or maybe it is happening right now but we do not realize it. Scary? Now this one is worse

AIs Are Teaching Themselves

AI1In the late 1990, Deep Blue, an AI designed to play chess, beat Gary Kasparov, the world chess champion in a chess game. In 2017 an AI called AlfaGo also beat a grandmaster in a Chinese chess Go which is said has hundred times more variations than chess. Also in 2017, Facebook has to shut down one of its AI because it started to ‘talk’ in totally new language that human cannot comprehend. Even before the shutdown, the AI showed defensive action in order to be kept alive.

AI learns from experience and all it does is problem solving. As a result, it gets smarter every time. Now humans just need to wait until it is smart enough to slip out of hands and break free, if not worse, conspire to overthrow humanity.…

Read more