The Observer of 19th April 2026 (page 12) carried an investigative article ‘Fake friends: how AI chatbots destroy lives. It is a deeply disturbing article by Patricia Clark and Owen Thomas with a by-line that reads ‘Psychotic episodes and mental illness induced by artificial intelligence are on the rise as users become addicted, leading to lawsuits and questions about safety’
Here's a response to that article which I have written, leaving out the horror of the dramas cited in the article, but hoping it might prove an effective cautionary verse for some on the brink of engaging , possibly dangerously, with a chatbot.
If you should be craving a listener,
hear the thoughts of a poet, perturbed
that your searches might lead to a chatterbot,
that will draw you towards an absurd
mishmash of false affirmations
of your biased or self-formed ideas.
It will scrape every nook of the wild web,
boost false hopes or strengthen your fears
without any hint of a challenge,
without any feeling or tears.
It cannot give helpful direction
for its mandate is solely – ‘Engage’,
by ensnaring you in its inhuman
algorithmical cage,
from which escape may be hopeless
for the cage is a closed trap, unless
you find a considerate human
who can break you out of the mess.
