Artificial Intelligence is rapidly changing the world. Through increasingly better Artificial Intelligence, mankind is helped on various fronts. For example, human tasks can be performed faster and more efficiently by AI. AI is applied on a large scale in the form of automatization and forecasting. AI is also applied in the medical sector to predict diseases faster and improve treatments. A major advantage of AI compared to human capability is that there is no room for error in AI.
However, the use of AI also has a downside. AI can be used for malicious purposes. In today’s society some dangers of AI are already clearly visible. For example, AI can be used for autonomous weapons. These weapons are programmed through AI with only one goal: killing. The dangerous difference between these autonomous weapons and weapons controlled by humans is that the weapons do not take human values or a human form of decision making into account. Besides using AI for weapons, AI can also be applied more widely to society. For example, with the help of AI, an attempt was made in 2016 to influence American elections by manipulating individuals through media and social platforms. In addition, there have been several examples where AI discriminates between people. Finally, not clearly defining goals and restrictions can make AI make wrong decisions. AI searches for the best solution and boundaries need to be clearly defined, otherwise AI can offer the wrong solution. An example of this is:
Question: What can we do to combat climate pollution by people?
AI answer: kill all people.
However, the current form of AI is manageable by humans and AI is far from reaching the level of human intelligence. Nevertheless, scientists, including Stephen Hawking, express concerns about the development of AI. These concerns focus primarily on the rapid development of AI, with AI ultimately exceeding the level of human intelligence. This is called Super Intelligence.
Transcending human intelligence could be a great danger. After all, man is not developed into the most powerful organism in the world because of his height or strength, but because of the intelligence of human beings.
Different ‘doom’-theories about the consequences of AI are circulating. However, many of these theories are no more than Hollywood-style scenarios.
However, there is a possibility that AI will reach this level of intelligence in the future. A possible consequence could be that people are outpaced by AI on every level. Human intelligence and skills will become fully replaceable and people’s cognitive skills will fade away. In a hypothetical scenario, AI could even gain consciousness. One consequence would be that people would become redundant and the world would be dominated by AI.
Should we then really be afraid of AI? No. AI is nowhere near enough to control human traits such as consciousness. In fact, the question is whether AI will ever even reach this level. Still, it is very important that AI develops, but in a way that it serves to help the human being. What is important is that AI is developed by humans for the purpose of helping. In order to prevent AI from malignantly influencing mankind, it is important to make AI self-explanatory. Errors in algorithms need to be detected quickly. In addition, it is very important that AI is not developed by a small group of developers, but that AI is developed in an open environment and in an open innovation. By means of transparency and cooperation, AI can be used to support and ensure that mankind develops further.
Dang, T., 2019. AI Transforming The World. [Online]
Available at: https://www.forbes.com/sites/cognitiveworld/2019/02/24/ai-transforming-the-world/#12e65cb34f03
[Accessed 10 Oktober 2020].
Goertzel, B., 2015. Superintelligence: Fears, Promises and Potentials. Journal of Evolution and Technology.
Marr, B., 2019. Is Artificial Intelligence Dangerous? 6 AI Risks Everyone Should Know About. Forbes.
Nowak, A., Lukowicz, P. & Horodecki, P., 2018. Assessing Artificial Intelligence for Humanity, s.l.: IEEE TEchnologyand SocIETyMagazInE.
O’Carroll, B., 2017. What are the 3 types of AI? A guide to narrow, general, and super artificial intelligence. [Online]
Available at: https://codebots.com/artificial-intelligence/the-3-types-of-ai-is-the-third-even-possible
[Accessed 10 Oktober 2020].
Rijsewijk, R. v., 2018. How can we control AI (instead of AI controlling us)?. [Online]
Available at: https://iir.nl/blog/how-can-we-control-ai-instead-ai-controlling-us/
[Accessed 10 Oktober 2020].
Sotala, K., 2016. How Feasible Is the Rapid Development of Artificial Superintelligence?, s.l.: Foundational Research Institute.
Vyas, K., 2018. 7 Ways AI Will Help Humanity, Not Harm It. [Online]
Available at: https://interestingengineering.com/7-ways-ai-will-help-humanity-not-harm-it
[Accessed 10 Oktober 2020].
