AI Assistants - Impact on Society
AI Assistants are slowly becoming an essential aspect of our lives. While not daily, I occasionally turn on the AI feature on my phone to observe the progress companies have made on their AIs. I usually ask them trivial questions followed by follow-ups to see if they can keep the conversation going. When AI assistants first came out, much like other people, I was also fascinated by their ability to share jokes, play games, and instantly search the web to give me accurate answers. However, over time, as my fascination with AI assistants died out, I stopped using them because I felt that doing the task independently was faster than trying to make the AI understand what I wanted. With time, as voice recognition and AI technology improve, I see more people handing over their trivial tasks to AI Assistants.
AI technology continues to evolve, offering enhanced interaction capabilities and efficiency in handling everyday tasks.
John Danaher, in his article proposes three arguments, I agree with the first but disagree with the rest.
The first argument is the Degeneration Argument, stating that you are likely to forget the ability to complete a task because you have given your responsibility to the AI.
This is true; the more you automate a task, the more likely you are to forget how that task was performed before automation. However, I disagree that this would be seen as a problem. The degeneration effect has been around since the evolution of humanity. For example, before the invention of wheels, people had to carry their items, offspring, and food over long distances by hand since this was necessary for survival. With the invention of the wheel and carts, people could carry items just as well, if not better, than they did without them. This resulted in fewer people carrying items by hand and more people using carts, negatively affecting their capability to take them by hand. Similarly, AI Assistants would negatively impact our ability to do trivial tasks; however, as humanity grows more advanced, the need for us to perform those trivial tasks would also reduce. AI Assistants’ impact is not something that has never happened before with the introduction of new technologies; it is a natural step towards progression.
The second argument is the Autonomy Argument, stating that we cannot take responsibility for tasks performed by an AI because we did not directly perform those tasks.
Understanding that the tools we use, like AI, extend our capacities and carry out tasks under our direction, helps us grasp how we retain responsibility for outcomes.
I disagree with this argument because no matter how small the task was compared to the task, it was performed due to our actions. The only reason an AI would do a task for us is that we have either told it to perform that task or programmed it to conclude that it needs to be completed. We have put an effort into accomplishing that task. For example, let’s consider autonomous robotic vacuum cleaners like Roomba. Even though you haven’t personally cleaned out the apartment, you have gone through the effort of purchasing the robot, charging it, and powering it on. This process itself puts you responsible for Roomba’s tasks. Using a Roomba is no different from using a specialized broomstick, considering it requires less effort and is much more efficient in accomplishing its task. Much like a Roomba, AI is a tool that makes it much more efficient to complete a task, thus putting the responsibility for its actions on us.
The last argument is the Interpersonal Communication Argument, stating that automating interpersonal messages would be deception; they would devalue the sentiments behind the message.
As AI begins to handle more interpersonal tasks, it’s important to establish clear boundaries and understandings about the role of technology in personal communications.
I disagree with this argument because, like the autonomy argument, if you set up the AI to perform a task, you have put in effort towards it performing that given task. If both the receiving and the sending parties knew about the technology being used to automate the interpersonal task, it would not be deception. I also feel that phrases like “I love you” hold sentimental value today because we know the feelings and emotions associated with the word. If an AI started using the phrase and it became common, people would simply move over to a new/different method to signify their feelings. I see this happening using irrational conversations that an AI would have difficulty recreating or through key conversations/phrases known only to the parties involved.
It is essential to consider that AI is just another tool we use to automate our day-to-day lives; much like any other tool, it will provide us with some benefits; however, it will also come with some disadvantages.