Artificial Intelligence and its Dark Sides
Besides all the advantages that artificial intelligence brings to our life we cannot surpass its disadvantages.
It is not necessary to go as far as Terminator movies…Danger is by our sides.
The concept of artificial intelligence is introduced to computer science by John McCarthy in the middle of the 50’s. It is a sub category of computer science that is basically used in such a way that machines are able to react to what they learn and see like real animals or human beings in addition of what they learned before. Because of that artificial intelligence is also called machine learning.
There are already many usage of machine learning that benefits our life even if we are not aware of it…
- When you surf in an e-commerce or movie web site the smart suggestions made are the exact example of artificial intelligence, especially new activities are made by users (see “recommendation engines” and “active learning”)
- When shopping with your bank’s credit card, withdrawing from an ATM or entering a bank free banking system machine learning can be used in fraud detection systems (see “Fraud detection systems”)
- Even the autopilot systems used on planes are artificial intelligence.
It is important to note that at this point, such positive systems and new generation technologies can be accessible by malicious people educated and financial resources as you are.
Many futurists, entrepreneurs and policy makers thinks that artificial intelligence is a nightmare like Terminators that can destroy people. Yet a lot of people like Elon Musk warns us about bots, robots or similar systems that can plan themselves, set a target, decide on this direction, open battles and bring the end of humanity.
Actually, the artificial intelligence we are currently in is a bit far. The term that can be used when talking about them is the term “general artificial intelligence,” or “general artificial intelligence,” in short, A.G.I. and even systems that are slightly similar to these types of systems are expected to exist after the earliest 2040s. So it may be early to worry about this…
On the other hand, the fact that our current AI capabilities are used even by a kind of cruel (“villain”) we encountered in the James Bond films frightens us enough. If you want to look at examples, evaluate yourself;
- Think for a moment about automatic cars that are being used in the near future and at least in test driving on US streets. Could it have been possible to program these tools for their own benefit, to load bombs into them, or just to “drag-and-drop AI” programmed for humans? Well, what tools do you have to block this?
- The harmful software, viruses and derivatives we encounter today are actually derived from rule-based code sequences that are not very long. Think for a moment, how did you get along with the malware that spreads by generating its own strategy, recognizes infected computer systems, and evolves on the road? If you do not have to worry about the lack of such systems yet you are wrong. “Stuxnet” The virus, which is used by the US intelligence units in 2010 for the Iranian nuclear program to last 9 months, is an example of this … Well, do you think Turkey has systems that can protect its nuclear power plants from such advanced “artificial intelligence” if not, how do we prevent our own nuclear plants from exploding on their own soil?
- And finally, if you think that most of your personal information, Facebook shares, blogs, and journal articles are open to the public, you can imagine how likely it is to develop an AI that writes and talks in your style, speaking of your interests. How convincing is the situation if you can convince enough AI that if you or your friends can access your bank account by posting “Phishing” letters, or if you can put your head in trouble by typing your word around, Think about how difficult it is to start …
My personal forecast for the near future is;
There will be a race of arms between the intelligent, intelligent and highly educated good children of the neighborhood and the intelligent and talented bad children of the same age. Almost none of the weapons here are weapons in the sense we know, but algorithms, learning systems, extremely powerful computers that can be used to teach them something, and extremely powerful math-based training systems. In other words, the PISA scores show the worst performance and we do not seem to be able to compete.
The facts are a bit further ahead, and mostly the fact that the bad guys are still successful with the old methods. But this fact is about to change rapidly …
Altan Atabarut, MSc.