AI Phone Bots: I Want to Hang Up Immediately

Passionate ranting about AI's role in society..

I have a deep hatred for phone AI bot systems. If you’ve ever called a Big Tech support line or even various retail stores, you’ve likely encountered an AI voice assistant trying to solve your problem before connecting you to a person. These systems are designed to reduce the load on support specialists; if they can resolve issues with a computer alone, they don’t need to hire as many support staff. The calculation is simple—they want to spend less money for the same quality of work.

This is understandable, but the way these systems have been implemented is completely wrong. Humans don’t view computers as equals, this is a very obvious statement. Even though computers have lately become more powerful and can imitate human interaction a lot better, the underlying reality remains: computers aren’t humans. They are (should be) assistants to humans, not replacements.

When I call a support line and some LLM bot greets me with,
"Hello, How's your day going?", I'M NOT GOING TO ANSWER.

AI bots aren’t people; they don’t care about my day. It is clearly a waste of time not only to ask the question, but to answer it. It is more confusing for the user than anything. Computers are tools we use to get information, access content, or complete tasks. That is the relationship, and it will always be that way, maybe until Robot slaves becomes unethical in 200 years.

If the people implementing these systems understood this, they could easily improve the dynamic of the interaction. If the AI started the conversation with a straightforward prompt like, "Hello. Please state your problem," instead of "Hello, how’s it going? What can I help you with today?", it would immediately establish the appropriate relationship and reduce confusion.

Hearing an automated voice use conversational language and phrases like "I," "me," or "us" just makes people want to hang up. It's clearly a deceptive attempt to imitate a real support specialist. The system is, in essence, trying to trick you.

Here’s what a better system would look like: it should start the conversation by clearly establishing that it’s a computer, there to provide initial support, and, if necessary, will connect you to a human who can fully resolve the issue. Rather than using conversational language, it should be straightforward and direct: "Please state your problem," or something similar. This approach makes it obvious from the start that you’re speaking with an AI bot, avoiding any confusion. There should never be a moment, 5 or 10 seconds in, when the user suddenly realizes they’re talking to a bot—it should be clear from the first statement.

Using phrasing like "Please state your problem" also simplifies things for the user, making it easy to ask straightforward questions like, "When do you close?" or "Do you have this item in stock?" Simple questions like these are easy for a properly configured phone system to answer—assuming, of course, that the store’s information is kept up to date.

Switching to this computer-assistant model would be incredibly effective. Not only would it answer more questions accurately, but it would also eliminate the irritation people feel when they expect to speak to a person and are instead greeted by an AI bot trying to sound human.

This model should be the standard, in my opinion. Human imitation in AI systems has always been irritating and generally useless. I can’t think of any real use case for a computer pretending to be a person. You might think of ChatGPT, but it's not really imitating a person. While it can respond in a conversational tone if asked, the reason it works so well is that people use it as a tool for information, summarization, and other general assistance. It fulfills the human-assistant role effectively, and it’s clear to users that its a computer on the other end, not a person.