Welcome to Neural’s Beginner’s Guide to AI. This long-running series is designed to give you a basic understanding of what AI is, what it can do, and how it works. In addition to the article you are reading, the guide includes articles on neural networks (published in order), computer vision, natural language processing, algorithms, artificial general intelligence, the difference between video game AI and real AI, and the difference between human and machine intelligence.
The most obvious solution to a particular problem is not always the best solution. For example: It would be much easier for us to throw all of our trash on our neighbors’ lawns and let them handle it. But it’s probably not the optimal solution for various reasons. At its core, such an action would be unethical because it forces someone to take your burdens in addition to their own.
Basically, it is unethical to pass the garbage on to the next person. And this is exactly what we need to focus on when trying to understand ethics in the field of artificial intelligence.
For the purposes of this article, when discussing the ethics of AI we are asking two simple questions:
- Is it ethical to build an AI for this specific purpose?
- Is it ethically correct to build an AI with these skills?
The first question concerns the intent of the developer or creator. Since there is no governing body to set the acceptable ethical standards we should impose on developers, we can do the best we can to determine the rationale for a particular AI system.
For example, if Google tells us that it has created an AI that can tag images in the wild, we accept its existence as a form of greater utility, assuming it was created without malice.
And thanks to this AI, we can type “puppy” into a search box on our phones, and Google will search our personal archive of thousands of pictures and display all of the pictures with puppies.
However, if you type “gorilla” into “Search” once and hit the “Pictures” tab, you will see pictures of black people. Regardless of the developer’s intent, they have created a system that upholds racist stereotypes on an unprecedented scale in human history.
The second question, “Is it ethically correct to build an AI with these skills?” Relates to the intent of potential external parties who may be inspired to abuse an AI system or develop their own.
For example, the development of an AI system that analyzes human emotions as recognizable in facial expressions is not inherently objectionable. An ethical application of this technology would be to create a system that notifies the driver when they appear to be asleep behind the wheel.
However, if you use it to determine whether an applicant is a good fit for your company, it is likely to be considered unethical. It is common knowledge that AI systems target white male faces. The systems clearly work better for one group than for another.
When it comes to ethical dilemmas, the popular situations that people like to talk about are rarely the ones that developers and developers actually face. Whether a driverless car decides to kill an old person or a group of children is not as common a problem as whether a database about people is so diverse that a system is robust enough to be useful.
Unfortunately, every entity in the modern world seems to have its own agenda and ethics when it comes to AI. The world’s superpower governments have ruled that autonomous killing machines are ethical, the general public has accepted deep counterfeiting, and the proliferation of mass surveillance technologies through devices ranging from doorbell cameras to the legal use of facial recognition systems by law enforcement agencies tells us it is is Wild West for AI in terms of ethics.
Published on February 26, 2021 – 20:30 UTC