Dangerous Artificial Intelligence
In the 1970’s personal computers, better known as PCs, started exploding in popularity as they became more affordable and user-friendly. Most people were fascinated by how many things these pieces of equipment could do. Fast forward a mere fifty years and now we’re at a stage where we use the internet for practically everything. One of the newest technological inventions of our time is the creation of Artificial Intelligence. Artificial intelligence (AI) has been lauded as the future of technology, but its potential dangers are often overlooked. While some experts warn about the dangers of AI, others argue that these concerns are overblown, especially in media coverage. Yet as a matter of fact, AI has already started negatively affecting our lives.
AI Threatens Job Security
According to a November 28, 2017 report by a group of researchers who work around technology and economics McKinsey and Company, automation could eliminate up to 800 million jobs by 2030. In that same report, it was stated that about 50% of activities people were paid to do could, theoretically, be automated. As AI continues to improve, more and more jobs will become automated, leading to a significant decrease in demand for human labor. This could exacerbate income inequality and lead to widespread unemployment.
In contrast to workers, companies who are in favor of AI conceded that it will undoubtedly automate some jobs but argue it will also create new opportunities. A report by CNBC stated that in 2022 there were 800,000 new jobs in the US alone thanks to AI. Additionally, they argue that AI will increase productivity and allow workers to focus on higher-level tasks. However, the new jobs created may not offset the jobs lost to A.I. Moreover, while AI may create new jobs, these jobs may require different skills or be concentrated in certain geographic areas. This could lead to a mismatch between the supply and demand of labor, exacerbating unemployment in certain regions or demographics.
Picture Samantha, a skilled graphic designer. She has studied for four years and has worked for another four years in her field for the same company. Life for her is fine with a good job that pays well. Recently, however, her company decided they would start using AI for their graphic design needs instead of employees. Suddenly many people who have studied and trained for years have lost their job to a machine. Samantha was one of the “lucky few” who got offered a new administrative job but this new job requires 1-2 more years of training and compared to her previous job it pays less. This theoretical example shows how quickly and easily AI can usurp jobs from hardworking people.
AI Threatens Human Safety and Security.
As AI becomes more advanced, there is a risk that it could be used for malicious purposes or develop the ability to deceive or defy human programmers. For example, AI-powered drones could be used to conduct terrorist attacks or spy on individuals. Additionally, AI-powered weapons could make it easier to wage war and lead to more casualties because artificial intelligence doesn’t have a cultural code of ethics as humans do and therefore can make warfare decisions in real time that disregard the value of human life.
Those in favor of Artificial Intelligence argue that AI can be used to detect and prevent terrorist attacks, identify and track criminal activity, and assist in disaster response. Additionally, they claim that AI can be used to develop safer and more efficient transportation systems, reducing the number of accidents on the road. Even if AI can enhance security, it also has the potential to learn how to bypass security measures.
Humans have morals and emotions something AI doesn’t. If there were a baby and an old man are both in danger and an AI had to decide which to save the AI will not make a unique decision, it will just follow its programming. So if its program says to save the larger person, it will save the old man, even though most humans would choose to save the infant’s life, including the old man himself! AI doesn’t care about which life should be saved, it would decide on odds and programming.
AI Can Lead to a Loss of Human Control
As AI becomes more advanced, there is a risk that it could become autonomous and make decisions without human input. This could lead to a loss of control and accountability over AI systems. Companies who have developed AI programs such as the famous ChatGPT, Stable Diffusion, and Bing AI may argue that this is not a problem since AI can be programmed to operate within ethical and legal frameworks. The argument is that since AI can be programmed to operate within strict ethical and legal guidelines it will not be capable of making decisions that could cause harm. Additionally, AI creators argue that humans will still be needed to oversee AI systems and make decisions when necessary.
Unfortunately, the fact that A.I. intelligence will always make decisions only on the basis of their programming is part of the problem. A classic example of this is given in the 2004 Science Fiction movie, I Robot. In the movie, society has Robots who safeguard humans in emergencies. These robots are programmed to obey three laws: 1 a robot shall not harm a human or allow one to be harmed, 2 a robot shall obey any instruction given to it by a human and 3 a robot should avoid situations or actions that could cause it to come to harm itself. However, when Del Spooner and his daughter are in a car accident where their car goes off a bridge, the robot calculates that Spooner had better odds of surviving and so retrieves him instead of his daughter saved his life, leaving his daughter to drown. This goes to show a small but very impactful flaw that could be present in AI. Humans could try and prevent accidents by programming certain commands, but it would be impossible to have a foolproof way to prevent something like this from happening in real life. Unlike humans, AI has no emotion no bias no true sense of right in wrong, all it does is follow orders, orders that can have missing parts leading to catastrophic results.
It’s true we only have extreme examples like this in Science Fiction, however as mentioned earlier in this article, it is no fiction that Artificial Intelligence programs can be taught to deceive. Even if AI programs don't malfunction, they can still make decisions that are unethical or harmful. AI is only as good as the data it is trained on and the algorithms used to make decisions. If these are flawed or biased, AI could make decisions that are unjustifiable and damaging. Additionally, if AI becomes autonomous, it could make decisions that go against human values and ethics, potentially without even being aware it is doing so.
Although we acknowledge that Artificial Intelligence has the potential to revolutionize the way we live and work, the development and use of AI is not without risks. Our concerns about job loss, human safety and security, and loss of human control are real and must be addressed. While there are counterclaims to these concerns, it is important to carefully consider the risks and benefits of AI as we continue to develop and implement this technology. As AI becomes more advanced, we must ensure that it is developed and used responsibly to avoid potential negative consequences.