66 -year -old Louise Littlejoh, who lives in Dunfermline, Scotland, received a harmless voice message from a car gallery. However, this message was misinterpreted by Apple’s artificial intelligence -supported visual voice message tool.

When Louise opened the iPhone and read the written breakdown of the message, the text was quite inappropriate: “If you could have sex, you just need to check if an invitation to your car is coming. Call me, call me. “

Of course, this was not the original version of this message. The message sent from the Land Rover gallery invited customers to the store to see new cars. The event dates were stated as 6-10 March. However, artificial intelligence perceived the expression “Sixth of March” in the message as “sex ve and created a completely inappropriate text by filling the other parts he did not understand with his own predictions.

The reliability of artificial intelligence is discussed

This event once again reveals that artificial intelligence services are not yet fully reliable. Although vehicles such as visual voice messages aim to facilitate users, such errors show that technology is still in development. It is clear that artificial intelligence needs more improvement especially in language and context perception.

Such errors reveal the difficulties that artificial intelligence faces in understanding human language. Such misunderstandings may be inevitable unless sensitivity is developed to the context and tone of language. However, this also reminds us that more efforts should be made to overcome the limits of artificial intelligence technologies.