In the 19th Century, a movement called Spiritualism swept the world. Spiritualists claimed they could communicate with the dead. They invited people to watch ghosts sending them messages by rapping on surfaces and moving objects around.

Spiritualism was a big hit. Major writers like Charles DickensThe Victorian novelist is credited with inventing Christmas as we know it today through the warm descriptions of it in his books. and Arthur Conan DoyleA 19th and 20th Century British writer, best known for the Sherlock Holmes series. were enthusiasts. Thomas EdisonAn American inventor who also pioneered sound recording and movie cameras. even tried to design a "spirit phone" that would allow people to record the ghostly messages for posterity.

But the Spiritualists were soon revealed to be frauds. They had created the spooky messages themselves with tricks and clever devices.

Today, few are convinced by ghost stories. But some think there are parallels with a much more modern phenomenon: the AI chatbotA chatbot is a computer programme that simulates human conversation or "chatter" through text or voice interactions..

Last week a new chatbot called ChatGPT made waves across the world. Experts are amazed at its long, fluent and intelligent answers.

Some think it could revolutionise our lives. Right now, if you want to do anything online, you have to navigate through a series of links to find the right website.

Chatbots can make all this much simpler. Whether we want to know the name of an actor in the film we are watching, set up a bank account or buy a Christmas present, the chatbot will be able to do everything for us.

Some fear the impact of chatbots could be more harmful. They warn it could be used to cheat on homework and exams. It could put millions of people out of a job.¹

But critics are not so sure. They think ChatGPT, like Spiritualism, is a parlour trickA trick performed to entertain guests. masquerading as a scientific breakthrough.

One of the first chatbots was developed in the 1960s by a scientist named Joseph Weizenbaum. It was called ELIZA, and it could only ask simple questions. It could not understand anything that was said to it.

But Weizenbaum was surprised to find its users were convinced it had human thoughts and emotions. Many began telling the machine intimate secrets, believing it understood their inner lives.

The more modern chatbots might be cleverer, but, sceptics say, they are just doing the same thing. We do not mind that the chatbot does not really understand what we tell it because we want to believe that on a deeper level, it understands who we are.

Yes: This could be the death of original writing. No-one will ever write anything for themselves any more, meaning that all writing will come to look exactly the same. We could be witnessing the end of an art form.

No: The chatbot is not nearly as sophisticated as it seems: it is just telling us what we want to hear. It does not have any real understanding.

Or... Even if the chatbot is not very sophisticated, it could still be dangerous. People might come to prefer talking to the chatbot than to other human beings, believing it understands them better than any human.