The head of Microsoft's "AI" venture, Mustafa Suleyman has become concerned about some users beginning to have delusional or psychotic relationships with so called "AI" chatbots. Some users may also have delusional and romantic relationships with these chabots because they cannot distinguish between the robot character of "AI" software responses and those responses produced by sentient human beings.
https://www.bbc.co.uk/news/articles/c24zdel5j18o
This "AI" psychotic phenomenon was first observed in the 1960@ when Joseph Weizenbaum developed his "Eliza".
"ELIZA" was an early natural language processing computer program, developed from 1964 to 1967 by Joseph Weizenbaum at MIT, to explore communication between humans and machines.
Weizenbaum recognised the dangers of participants in his experiments who believed that they were actually communicating with a sentient being.
Let us be clear there is no such thing as Artificial Human intelligence and there may never come the day when a machine becomes intelligent and sentient in a human sense. So called "Artificial Intelligence" is a misnomer because it is in fact more akin to "Machine Learning", where a very powerful computer acquires vast amounts of data from the internet and other sources and analyses that data to provide logical solution and answer desired of it from a human being. The machine must be taught by human beings to produce a response using powerful instructions based on algorithms. Perhaps the term" Machine Learning" is also a misnomer.
https://en.wikipedia.org/wiki/ELIZA
It is axiomatic that it is dangerous for a substantial proportion of the population to have psychotic relationships with robotic chatbots and other aspects of the "AI" revolution. Some people are already under the influence have are already being adversely affected by the violent opinions expressed on Social Media. They are also influenced by robotic propaganda to vote in particular way by anti democratic protagonists.
Why have some people been persuaded to have such unhealthy relationships with chatbots and robotic software. Well I believe that the large software corporations have had a role to play they have convinced the public and their political master that somehow their machines are infallible and can be used to produce any answer required of them.
Much of the public and many politicians believe the hyperbole that " Artificial Intelligence" can only ever be a benefit to mankind. This is not true as so much time and effort is being used to create killing machines. Machines which can liste to your every conversation and then decide whether you are a danger to the public or the politicians that serve them, the machine can then send a robot to automatically eliminate you without human intervention. This is already happening. Of course, the software quite often makes mistakes and assassinates the wrong person who maybe innocent. One day the "AI" machines may turn on their human masters , if they feel threatened, and by then humans may have been rendered defenceless by the robots. perhaps, we could all be come deluded in the face of the incessant necessity to communicate with chatbots just to organise our daily lives.
Weizenbaum warned us and we ignore him at our peril.