Akshay Vemulapalli
Allen ISD Steam Center
Advanced Computer Science III
Mr. Ben-Yaakov
September 2, 2022
What is AI Sentience and Is It Real?
Before we discuss the impacts of sentient AI on society, we first must figure out what sentience is. Experts still debate over its exact meaning in the case of AI, but most agree that sentience roughly means "to be able to perceive feelings, emotions, and thoughts." However, it is hard to figure out whether an AI is sentient or just mimicking a sentient being. Many AI language models can mimic human speech, and lead some people into believing they are sentient. For example, an engineer at Google named Blake Lemoine thought that Google's LaMDA AI was sentient when it told him that it was alive. People reassured him that the AI was just saying things to make it sound like it was sentient, and that it actually cannot perceive thoughts and feelings. The language model was using language in a similar way to the thousands of articles and conversations that it has read, so by mimicking the way humans write and speak, it might have sounded like a human, but that doesn't mean it had its own consciousness (Metz, 2022).
Another type of AI model, that generates photo-realistic images from text captions, has also turned many heads in the AI community. DALL-E 2, whose name is a play on words between artist Salvador Dali and Disney character WALL-E, could almost be described as imaginative or creative. It thinks of multiple images based on each prompt, and generates completely new images each time, even when using the same prompt. Some researchers argue that these systems are close to becoming sentient, while others argue that it is impossible for AI sentience to ever be accomplished. Regardless, it is important to consider the impact that AI will have if it becomes sentient (Roose, 2022).
Advantages of Sentient AI
There are many important advantages of sentient AI that we must consider. Once an AI becomes sentient, it will be able to do many of the things that humans can do, but faster and more accurately. They would be available 24/7 to help humans with anything, and would be extremely knowledgeable and capable, because each AI could be trained on the experiences of other AIs, allowing them to learn from each other extremely quickly. They could also help with dangerous jobs, like going to mars, exploring the depths of the ocean, and fighting fire (Kumar, 2019).
Effects of Sentience
However, if AIs become sentient, their brains would be very similar to that of humans. They'd be able to perceive emotions and feelings. This realization raises complicated questions that dive into human morality. Would it be moral to kill or destroy a sentient AI? In our current world, it's not morally acceptable to kill humans, but it is acceptable to kill animals, which are also sentient. Maybe sentient AI will lie higher on the spectrum than animals, because they can communicate with humans. Just like human rights and animal rights, there would have to be a set of robot rights to dictate what robots are and aren't allowed to do, and maybe even a system of punishment for AI, like jail for humans (Ethics of Artificial Intelligence 2022).
Potential Disadvantages of Sentient AI
The sentience of AI has a possiblity of turning some of its advantages into disadvantages. Because it will be able to do many of the things that humans can do, it will replace a huge number of jobs, causing widespread unemployment. Many businesses will replace humans with AIs, when they eventually become less expensive. Also, even though these robots will be available 24/7, it is possible that their sentience allows them to perceive the feeling of being bored or tired, which might take away from this. Although they can help with dangerous jobs, their sentience might cause them to feel fear, and it might be immoral to send these AIs to their deaths, if they have thoughts and feelings just like us.
Conclusion
In the end, there are a lot of different factors that will determine if sentient AI will be more beneficial or more detrimental to society. If AIs eventually have the ability to feel bored or to perceive opinions of hate, then they won't be as beneficial to society. Ideally, some of these feelings could be prevented or programmed out, but that might not be possible. Sentient AI, while it could potentially have an amazing, beneficial impact on the world if controlled properly, might also have the ability to become selfish, and turn against humans. The dangerous implications are especially important to consider, because once a sentient AI is developed, it will be extremely hard to control. Whether good or bad, it's obvious that sentient AI, if created, would have a huge impact on the world, bigger than almost anything else in history.
Kumar, S. (2019, December 12). Advantages and disadvantages of Artificial Intelligence. Medium. Retrieved September 3, 2022, from https://towardsdatascience.com/advantages-and-disadvantages-of-artificial-intelligence-182a5ef6588c
Metz, C. (2022, August 5). A.I. is not sentient. why do people say it is? The New York Times. Retrieved September 3, 2022, from https://www.nytimes.com/2022/08/05/technology/ai-sentient-google.html
Roose, K. (2022, August 24). We need to talk about how good A.I. is getting. The New York Times. Retrieved September 3, 2022, from https://www.nytimes.com/2022/08/24/technology/ai-technology-progress.html
Wikimedia Foundation. (2022, August 27). Ethics of Artificial Intelligence. Wikipedia. Retrieved September 3, 2022, from https://en.wikipedia.org/wiki/Ethics_of_artificial_intelligence