Although the chances of AI driven technologies are decades away from developing true human responses and consciousness, the people behind such technology still need to consider the ethical implications. This is not to say that one day humanity will be enslaved by the technology we created. The importance is in the inherent biases the people behind the technology still have. Ethical AI is something that needs constant attention as the digital world evolves, but what does that mean exactly?
What exactly is ethics in AI
Ethics in AI, also called robot ethics, comes down to the algorithms that create true artificial intelligence. The key here is the person behind these algorithms that make AI systems. If an algorithm hasn’t paid attention to certain details, the algorithm has the chance to produce the wrong results. As algorithms become more technically complex, one of the most important aspects to consider is the human imprint on said algorithms. Bias in AI always goes back to the choices the developers and researchers make. This is not to say that they are ethically flawed, but the way that they present data to the machines could produce ethically ambiguous results.
Take the example of a simple chat bot on a website. If the people behind the natural language processing of the chat bot all spoke the same language and came from the same backgrounds, the chat bot would inherently be more inclined to ‘think’ like the people who developed it. Problems could arise if someone from a different background were to use the technology, they could have issues, as the chat bot has not been trained to consider other backgrounds or lexical methods of communication. This is not the problem of the AI itself, but the fundamental human experience that has been narrowed by the tight parameters of the AI algorithms.
Why we need to consider ethics in AI
Relying more on AI in everyday situations means that the algorithms underpinning AI must have human welfare at its forefront. As we enter the world of AI making decisions alone and freely, it becomes more critical that AI can protect us rather than destroy us. If we are to trust AI and the digital landscape, it must work alongside humans. Often, in robot ethics, the focus is on the resulting actions of the AI. But the developers and their inherent biases must also be morally judged, as they are the ‘brain’ behind the AI technology. This issue is often considered when war and technology interlink. It must be intrinsically clear to an AI that the harm of humans is against all programming.
But the ethics of AI goes beyond the developers that write the algorithms. It is not just a moral question of right and wrong. There is debate over the use of AI in human roles such as therapy, social work and even policing. This is where it gets ambiguous and questions such as empathy come into the equation. Can job roles that rely on empathy be given to AI? Can we teach the creativity of empathy to robots and AI? And if so, could that have detrimental consequences? A huge chunk of empathy originates in the experiences that a person goes through in their lives, it comes from intuition. These are two traits that take years to develop and trust in a human. So, the biggest question remains, do we have enough data to teach this to AI?