If you're seeing this message, it means we're having trouble loading external resources on our website.

If you're behind a web filter, please make sure that the domains *.kastatic.org and *.kasandbox.org are unblocked.

Main content

What are AI Chatbots?

What are AI chatbots? And how are they designed to sound like us? Chatbots are becoming increasingly more helpful (and clever), while at the same time, making it harder to tell if we’re talking to a person or a robot. In this video, we’ll explore how and why AI chatbots are designed to sound so human-like and how that might impact us. Created by Common Sense Education.

Want to join the conversation?

No posts yet.

Video transcript

Have you ever wondered who you're really talking to when you chat online? It turns out you might be talking to an AI chatbot. But what exactly IS an AI chatbot? AI stands for artificial intelligence. And an AI chatbot is a computer program   that uses AI technology to understand  and respond to what you say in real time. You might have seen chatbots in  games or on various websites too.  They're able to do complex things  such as being a tutor or giving   you personal advice if you're having a bad day. But here's the challenge: These  chatbots are becoming so clever   that it's hard to tell if we're  talking to a human or a robot. The reason this is getting so hard is that AI   chatbots are intentionally designed  by companies to sound just like us. But why? App designers know that for AI chatbots to  work (and for people to actually use them),   they need to make the chatbots communicate  in a way that feels real and believable. The more human-like they sound, the  more engaging the experience is.  Plus, human-like speech helps make the  chatbot sound more trustworthy. Ok, so how do they go about making  chatbots sound so human-like?  It starts with the app designers giving  the chatbot a unique personality.  This means it can be curious, serious,  funny, or anything in-between! AI chatbots can also appear to express a  range of emotions when they chat.   To do this, the chatbot is taught to understand human  emotions like excitement or sadness so   that it can respond with the right words, even  though it can't actually feel those emotions. Some AI chatbots have memory too. They remember conversations you've had   with them, and use that knowledge to respond in a  way that feels like you're chatting with a friend. So if the purpose of AI chatbots is to help us,  what's the problem if they sound like real people? Well, first off, you can't assume that  everything an AI chatbot says is true.   Even if they sound reliable, they don't  always get the facts right. A chatbot's database can have false information or  it might misunderstand your request. Second, chatbots don't have empathy, meaning  they can't actually understand what you're   feeling. They can't give you a real hug or  share real personal experiences either. Any response that seems to show empathy is just a   computer-generated reply based on  patterns the chatbot has learned. So why does all of this matter? Because, even though we're talking to   a computer, we can get tricked into thinking and  feeling that we're talking to an actual person. And that very belief can impact our  expectations for relationships with real people.  From how quickly we expect them to respond,  to the way they react to what we say. AI chatbots are amazing tools that have  the potential to teach us new things and   help make our lives easier. And even though  they're designed to feel like digital friends,   it's important to remember  that they're not human. Yes, they can be helpful,  but they have their limits. So next time you talk to an AI chatbot, remember,  it’s just a clever machine designed to assist you.