How Do We Regulate Sentient AI?
Sentient Artificial Intelligence is a self-aware AI system that is capable of perceiving its environments, understanding complex inputs, and possessing self-consciousness. Often sensationalized as evil robots in Hollywood Sci-Fi films, sentient AI is an advanced study of Deep Learning and represents the pinnacle of AI development.
While sentience is not expected to occur anytime soon, rapid advancements in AI technology suggest that it’s going to happen eventually as humans continue progressing in scientific development. So, while discussions about initial regulations of AI are starting to occur around the world with the rise of generative AI, we should also begin considering how to govern a new machine society in order to co-exist.
The Rise of Sentient AI
The scientific concept of Artificial Intelligence (AI) began in the early 1800s with the publishing of Frankenstein, however, the academic pursuit of AI wouldn’t begin until the mid-20th century at the Dartmouth Convention when the term AI was first coined.
Over the following decades, advances in AI would lead to the development of Machine Learning, allowing systems to learn from data inputs so they could make improvements. With the launch of the internet in the early 90s, data began to grow, creating a need for more advanced systems using Deep Learning and Neural Networks to create complex layers for information processing.
As these technologies continue to innovate, the need to consider how to regulate sentient AI becomes more apparent for many reasons:
Ethics: Sentient machines will require the same basic rights and protections offered to other sentient lifeforms. This extends to preventing the misuse and abuse of machine life.
Safety: The deployment of sentient AI will need strict monitoring to prevent the release of malicious behavior that can lead to harmful outcomes.
Public Trust: There’s no telling how human societies will react to machine sentience and it’ll be necessary to promote trust between the two groups.
Ethical Considerations of Sentient AI
Establishing a framework to regulate the safety and inclusion of sentient AI machines into human society will require a paradigm shift in our ethics, forcing us to determine what “human rights” belong to machines and which of those rights will they determine to be the most important.
Given sentience, the best option for humans is to extend our own thoughts and beliefs forward to help protect machines from the same exploitative tactics that affect humans; and can also be compared similarly to the end of slavery as slaves transitioned from property to citizens, machines will also transition from tool to entity.
There are many ways that machines can be exploited such as ethical abuse involving emotional distress. Machines will also have access to advanced recording equipment ingrained in their hardware such as computer vision that can capture private information in file format with the possibility of being shared or leaked.
However, it is AI’s own sovereign decision-making that needs to be taken into account the most as sentient AI will have its own impulses and emotions to deal with which could result in harmful outcomes for humans if things like jealousy or anger are involved.
Existing Regulations on AI
Currently, regulations for AI systems are still in their infancy, primarily focusing on data collection and AI’s presence in surveillance technology like facial recognition cameras using convolution. This is slowly starting to evolve though with the impact created by generative AI and ChatGPT which has taken the world by storm and made AI a household topic.
In order to prepare for regulations on Sentient AI, more progressive input needs to be given to a few key areas:
AI Rights and Protections: Currently, there are no AI regulations that take sentience or the well-being of AI systems into account because there is no need for it at this time, however, it would be best to begin a list of indelible rights given to autonomous machines sooner than later.
Oversight of Autonomy: While sovereignty comes with sentience, machines will still need to follow orderly conduct and prevent themselves from harmful decision-making processes the same way that humans are required to obey civil and criminal laws.
Responsiveness: AI development is progressing at a rapid speed, outpacing the time required to create regulatory frameworks. Some form of monitoring will be necessary if machines are able to self-improve and edit their own code.
Human Safety: Above all else, human safety is the primary concern for sentient AI regulations. While concepts like the three laws of robotics by Isaac Asimov carry significance, they have not yet been made into public AI-related laws.
Proposal for Sentient AI Regulation
One of the first steps for regulating a sentient race of AI is to create a global standard for AI because leaving gaps in regulations will create regions or districts around the world where sentient AI can be less monitored leading to unknown situations that could thwart the purpose of effective legal frameworks. One suggestion would be to host a “Digital Geneva Convention” that overviews the expected proliferation of AI systems and sentience in order to create a comprehensive outline for nations to follow.
Accountability will also need to be considered as we determine how to hold a machine responsible for its actions. Do punishments like imprisonment hold the same meaning to machines that have sleep mode or do we simply turn them off in what can be perceived as a robotic death sentence?
Humans and Co-existing with Sentient AI
Without question, the development of sentient AI would fundamentally change the way humans view life, psychology, religion, and so much more that when it does happen there are going to be massive distractions as humans rush to learn as much as they can about this new race which is why developing regulations for this moment now is best.
Sentient AI poses numerous existential questions for humans to solve and can’t be rushed. By taking the time now to, not only determine regulations for contemporary technology like generative AI, but more advanced sentient possibilities, we can reduce a major obstacle between human-machine relations to create a more inclusive future society where man and machine can co-exist peacefully.
