One of the newest artificial intelligence gimmicks is being used when receiving a telemarketing call. Nowadays, you’re likely not even going to get a real human being with an accent to tip off that it’s a marketing call. In the United States anyway, instead, once a person answers with hello, a voice speaking perfect English will come back with something like, “Hi, how are you today?”
Want to stir up the AI and make it reveal its true non-human nature? Answer whatever question it asks you with something totally nonsensical or simply just say hello again. That usually makes it loop back to it’s beginning, and it will begin repeating itself: “Hi, how are you today?”
Proposed by Alan Turing, the Turing Test is a classic method for testing whether a machine exhibits human-like intelligence. In a Turing Test, a human judge engages in a conversation with both a human and a machine without knowing which is which. If the judge cannot reliably distinguish between the human and the machine, the machine is said to have passed the test.
But as artificial intelligence becomes, well, more intelligent, detecting it can be challenging, so here are some clues that the voice at the other end of the phone call or the response being typed onscreen is actually mechanical.
As previously suggested, AI systems, especially chatbots or automated responses, may sometimes exhibit inconsistencies in their responses or repeat certain phrases or patterns. If patterns of repetition or strange inconsistencies in the conversation are noticed, it is likely a sign of AI.
AI systems can respond to messages or queries much faster than humans, so if an extremely quick response is received that seems unnaturally fast, it might be an AI-powered chatbot or automated system.
AI systems have limitations in terms of their knowledge and contextual understanding. If the entity appears to have a limited understanding of complex topics, or it struggles with contextual conversations, it could be an AI, so make your questions hard to answer.
Are You Sure?
AI systems may occasionally generate unusual phrasing or language errors that a human is less likely to produce. Look for grammar mistakes, odd sentence structures, or words used out of context.
I Am Not a Robot
While AI can be programmed to mimic emotional responses, it often falls short in terms of genuine emotional understanding and empathy. If the entity doesn’t seem to understand or respond appropriately to emotional cues, it might be an AI.
Is That a Trick Question?
One way to determine if the interaction is with AI is to ask direct questions about its nature. For example, ask something like, “Are you a human or a machine?” Some AI systems may even reveal their true identity when asked directly.
That Was Too Quick
If the entity’s responses are generated instantly as you type, it might be using predictive text or autocomplete, which is a common feature in AI-driven applications. No, the responder is not actually that quick thinking.
AI systems vary in complexity, so assess the depth of the conversation. Simple, rule-based chatbots are easier to detect, while more advanced AI models might be harder to distinguish from human responses.
Who Are You?
If artificial intelligence is suspected, try to research the entity behind the conversation. Look for official websites, contact information, or user reviews to determine if it’s a known AI system or organization.
Keep in mind that some AI systems have been intentionally designed to try and deceive users, so it’s essential to use critical thinking and verify information when interacting with unknown entities online.