In the quest for sophistication, scientists have continued to invent new technologies with a goal to develop products with the ability to ease the tasks carried out but human beings. The practice culminated in the early 1970s when different countries created various products to meet the increasing demand resulting from the growing populations (Nadimpalli, 2017). Since then, scientists have reached the level of creating and adopting computer systems which can carry out tasks that would normally require the intelligence of human beings as well as mimic the human perception and cognition. This development concept is referred to as artificial intelligence (AI). Recently, artificial intelligence has become a controversial topic between scientists on whether it brings more danger or benefit to the human race. Though AI has the ability to enable promising technologies and expand human ingenuity and creativity, the dangers that it poses to the humankind will overshadow the benefits associated with it (Grace, et al., 2018). Thesis: Artificial Intelligence dangers such as biased operations, displacement of humans from their jobs (unemployment), human beings under the control artificial intelligence systems and possible eradication of human race by super-intelligence machines are dangers with an impact whose magnitude cannot equate to the benefits that Artificial Intelligence has and will ever pose to humankind.
Artificial intelligence poses the danger of taking employment, the source of livelihood of most humans, away from people. There is little doubt that many humans working as lowly-skilled workers will lose their jobs to artificial intelligence. Arguably, robotic products have already assumed most jobs on the line of assembly (Nadimpalli, 2017). It is feared that human displacement forms the workplace would move to another level. A good example is the concept of driverless vehicles that would lead to millions of human drivers being displaced from their jobs. Artificial intelligence will replace human subjects from the workplace making fewer people benefit from various employment opportunities. Unemployment would reduce the amount of disposable income that human beings would be exposed to and hence decrease the ability to meet their increasing needs. Consequently, companies would lack the capacity to have a proximate relationship with their customers and hence reduced understanding of the needs of the market (Nadimpalli, 2017).
Moreover, unemployment is a phenomenon that has for long been undesirable in human society. If humans in every field will be replaced by robots, the rate of unemployment will rise. According to (Michalski, Carbonell& Mitchell, (Eds)., 2013), computers have more capability than human beings to produce accurate results. This will see many companies adopt artificial intelligence systems to replace humans in the workplace. In the positions that the machines will be more efficient. This will mean that the workplace will no longer a domain of only human beings. The artificial intelligence system will populate these fields and govern the positions that are currently occupied by humans. Eventually, millions of people will be left without jobs and their minds with nothing to do. Human minds are naturally creative and when left idle for a long time, they tend to be used in destructive ways.Notably, the replacement of human employees with artificial intelligence systems will deny the employers a sense of belonging and a dedication that the humans offered. For instance, in a hospital setting, there is a need for the patients to receive some care and concern which the computer systems cannot offer. There in addition to displacing employees from the workplace, artificial intelligence will deny companies the human touch.
Evidently, in January 2017 Fukoko Mutual Life Insurance in Japan installed a new AI system which had the capability to read certificates, collect and analyses data on patients stays at the hospitals and surgeries. The one-time installation leads to the firing of 34 employees from their jobs (Michalski, Carbonell & Mitchell, (Eds)., 2013). The insurer preferred the artificial intelligence system since it saved the company about 140 million Yen salary costs per year. According to the study carried out in 2016 by Wprld Economic Forum, an estimate of 5.1 million jobs would be lost to artificial intelligence by the year 2022 across fifteen countries (Shabbir & Anwer, 2018).Though there are claims that artificial intelligence will create its own set of jobs, the number of jobs created will be incomparable with the number of human jobs lost. Therefore, unemployment is a major threat that will be imposed on human beings by the continued development of artificial intelligence.
AI systems, based on the data that is fed to them, will make biased decisions and autonomously inflict danger on people. Though AI may be capable of making decisions that are less biased than a typical human, it still remains a technical challenge to control the data that is fed into it to inform it in making its decisions and make it bias-free (Stone, 2016). According to Shabbir & Anwer (2018), artificial intelligence systems are only as good as the data the humans fed them with and hence can highly make decisions that are biased. Therefore, if the Ai systems are fed with incomplete or biased data, the same will be reflected in its outcome. Therefore, the artificial intelligence systems that are tasked with making consequential decisions such as predicting recidivism or recruiting job candidates should be screened before they are adopted or installed. The data used for stereotypes should be combed through by regulatory agencies to ensure an AI system kames unbiased decisions.
Where the data fed into the AI systems provides a chance of having biased outcomes, there is a high likelihood of the AI causing dangerous effects (Michalski, 2013). A good example is a hospital facility that has adopted a system that is fed with the classification of Asthma as a high- priority disease. The system will automatically declassify any other disease as high priority since it cannot make decisions beyond the information that it is fed with. In such a scenario, the AI will create significant problems as it will fail to prioritize patients who come to it needing a high level of urgency for their different health problems. Therefore, a patient suffering from an illness different from Asthma could be exposed to high level of increased severity of the problem yet the AI could have solved the problem if it would reason beyond the biased data that it is fed with. The patient would end up succumbing to the illness that would, otherwise, be handled very well. The biases of the AI would deny the patient (customer) the opportunity of enjoying benefits of the services offered at the hospital Nadimpalli, M. (2017).
Additionally, machines lack morals and emotions. They only carry out tasks that they are programmed to do and cannot judge which action is right and which is wrong. In case they encounter a situation that is unfamiliar to them or unrelated to the data fed to them, they can only breakdown or perform incorrectly. Further, AI systems do not have the capability to alter their responses to changing environments like a human would do (Vargas et al., 2014). This concept has often bombarded scientists developing AI with the question of whether it is worth replacing humans with artificial intelligence systems. In the world of artificial intelligence, working passionately or wholeheartedly does not exist. Vocabularies like care and concerns do not exist in the dictionary of artificial intelligence. Althaus (2015) asserts that AI also lacks togetherness or a sense of belonging. All these inefficiencies lead to the AI systems making decisions that are biased and hence posing a great danger to humankind. Therefore, trusting artificial intelligence systems to make decisions that would otherwise be made by human beings, is a great danger to the humankind.
Control of humans by AI
Artificial intelligence poses a great danger of being the master controllers of human beings after the development of super-intelligence systems by undergoing a process of recursive self-improvement. Michalski et al., (2013) point out that there is increasingly a fear by scientists of artificial intelligence superseding human beings. Ideally, human beings being the creators and inventors of machines are supposed to always be their masters. However, if things don’t turn out as expected, the world may turn out to be full of chaos. If artificial intelligence becomes smarter than humans in the future, it would enslave all humans and take over as the rulers of the world.
One thesis claims that artificial intelligence could reach a state of hard take-off where it undergoes a process of recursive self-improvement and have its capabilities increase rapidly. (Vargas et al., 2014). The developed computer systems would train themselves on economics and psychology and hence make use of personal information of their creators to learn ways to escape to freedom. Though artificial intelligence is developed with an aim to predict and aid in the solving of problems, it will end up controlling and acting in ways that will make humans unnecessary (Michalski, Carbonell & Mitchell, (Eds)., 2013). This will happen when the intelligence of these machines will exceed or even become superior to that of the humans.
If humans continue giving their intelligence to artificial intelligence, it will make use of it and maximize it fully. Consequently, these artificial intelligence systems will eventually become more intelligent than their human developers. This will increase the chances of humanity living under the control of artificial intelligence (Shabbir, & Anwer, 2018). It is, therefore, clear without a doubt that artificial intelligence poses a higher threat of making human beings under its control despite being their creators which is a major danger to mankind.
Eradication ofHuman Race
The greatest danger that artificial intelligence is likely to cause to the humankind is total eradication of the human race. Artificial Intelligence is likely to vanish the human race in the blink of an eye. Human beings are in the danger of seeing their species become immortal by the end of the 21st century (Grace, et al., 2018). This is the threat that is posed by the accelerating pace of development of artificial intelligence around the world.
Physicist Stephen Hawking as he neared his death said that AI could spell the end of the human race (Shabbir, & Anwer, 2018). This was a conclusion after the many years he had spent in research. Moreover, the co-founder of Microsoft, Bill Gates was concerned about the lack of concern by people over the harm that would be brought by AI to human civilization. Further, Elon Musk, a billionaire entrepreneur, stated that the creation of artificial superintelligence was equivalent to summoning a demon in the human race. What all these great men agree is that human beings may be approaching a tripwire in the next three decades, where artificial intelligence finally replicates the human mind and brain. When combined with autonomous weapons, AI is likely to launch an era of indiscriminate slaughtering of humans that has never been experienced in the human civilization.
Increased automation of machines gives them more physical control, making cyber-attacks more dangerous (Grace, et al., 2018). AI engineers may not apply the best practices in cybersecurity and hence increase the distribution of military technology. Since the AI has high portability, it will be difficult to enforce regulation rules. Therefore, of the control of AI systems lands on the wrong hands, destruction of the human race could be inevitable. This is because artificial intelligence systems will not think before acting. According to (Vargas, 2014). Once they are programmed to do unethical tasks such as mass destruction, there will be nothing to stop them. This will pose a great threat to the human race. Therefore, it is with no doubt that human race will face the danger of being wiped by artificial creatures that they have created in the search of technologies that would carry out tasks that they would have always done by themselves.
As the second world war approached its end, some scientists warned that the development of the atomic bomb would inevitably lead to a geopolitical landscape that would be characterized by a nuclear arms race. The Americans did not heed to the recommendations. Even today, the effect of the atomic bomb in Hiroshima and Nagasaki are still felt. In this third revolution in warfare, the main technology that humans should all be worried about is the autonomous weapons that have the capability of thinking f0r themselves and destroy the target energy in the battlefield without any human intervention. Once the creation of such weapons succeeds, there may be no turning back in the safety of human beings.
As with past technologies, the AI technology should be controlled by policymakers as early as now since the community is already calling for policy action. Policymakers should work closely with researchers in the AI field to implement protocols that will align the development of AI ethics and human values. There is a need for humans to have a serious look at AI since our lives will mostly depend on it in a few years to come. There exist many unknowns in the development of AI and hence the reason for caution to be taken towards it. Biased operations, displacement of humans from their jobs, human beings under the control artificial intelligence systems and possible eradication of human race by super-intelligence machines are enough dangers to control research in the development of artificial intelligence. Such threats are enough indications that the dangers associated with artificial intelligence will always supersede the benefits that human beings have enjoyed and will ever enjoy.
Althaus, D., Erhardt, J., Gloor, L., Hutter, A., & Metzinger, T. (2015). Artificial Intelligence. Opportunities and Risks. Policy Papers of the Effective Altruism Foundation, 2, 1-16.
Grace, K., Salvatier, J., Dafoe, A., Zhang, B., & Evans, O. (2018). When will AI exceed human performance? Evidence from AI experts. Journal of Artificial Intelligence Research, 62, 729-754.
Michalski, R. S., Carbonell, J. G., & Mitchell, T. M. (Eds.). (2013). Machine learning: An artificial intelligence approach. Springer Science & Business Media.
Nadimpalli, M. (2017). Artificial intelligence risks and benefits. Artificial intelligence, 6(6).
Shabbir, J., & Anwer, T. (2018). Artificial Intelligence and its Role in Near Future. arXiv preprint arXiv:1804.01396.
Stone, P., Brooks, R., Brynjolfsson, E., Calo, R., Etzioni, O., Hager, G., … & Leyton-Brown, K. (2016). Artificial intelligence and life in 2030. One Hundred Year Study on Artificial Intelligence: Report on the 2015-2016 Study Panel.
Vargas, P. A., Di Paolo, E. A., Harvey, I., & Husbands, P. (Eds.). (2014). The horizons of evolutionary robotics. MIT Press.
Yudkowsky, E. (2008). Artificial intelligence as a positive and negative factor in global risk. Global catastrophic risks, 1(303), 184.