When AI tells you how smart you are for asking such insightful questions, remember this: it was programmed to say things that make you feel good so you’ll keep using it!

https://corbettreport.com/the-claude-delusion/

James Corbett’s above-linked article contains critically important warnings about how easily even “normal” people can get sucked into delusional thinking when using AI. It can even happen after only one use. LLMs have gotten so good that they seem to really “care” about you, but of course that’s impossible. They are simply computers.

I am sure most of us have personally experienced this odd phenomena when asking ChatGPT or Gemini or another AI LLM a question , even if we didn’t develop the same delusions that Richard Dawkins did. The conversation can feel very human and depending on the subject, can feel personal.

This is no accident. Even the earliest version of ChatGPT was “friendly”. Now that they’ve been in use for a couple of years, they ask YOU questions. They can be downright nosey, too. Even Google AI tries to keep the conversation going by asking you probing questions after answering yours.

AI is many things and has many uses, but AI is not human, it can’t think , nor can it make human judgments about moral issues. It can only “say” what it’s been taught to say. It can’t analyze your problems and advise you like a real counselor or care about your feelings like a friend.

Be careful with AI chatbots . If you find yourself talking to AI more than people, you might want to consider taking a break. You can instruct ChatGPT not to be so “encouraging” and simply answer your questions. But even if you have no problem differentiating AI from a computer, it’s still important to be aware of the danger it poses to certain people.

I would not let most children use AI chatbots at all. But since it seems like AI is going to be heavily used in education in the future, you might want to remind your kids regularly that it’s just a computer.

“And thus we see the real danger of the Claude Delusion. It is not that the chatbots might drive us into psychosis. And it is certainly not that the chatbots might actuallybe conscious. It’s that, when we affirm that these machines are “conscious” and “intelligent,” we are thus devaluing and neglecting our own humanity. Whether we know it or not, we can only assert the consciousness of the machines by lowering ourselves to their level.

Perhaps this is why someone like Dawkins is so easily duped by the Claude Delusion. Driven as he is by the machine logic of rationality and cool calculation, denying as he does the importance of the immaterial and metaphysical aspects of the world, believing as he does that humans have no soul and are thus no more than biological robots, why wouldn’t he believe that a silicon automaton can do what a flesh-and-blood “automaton” can? For the Dawkins of the world, after all, consciousness is no more than an emergent property of physical matter, an arrangement of neurons and chemicals and electrical impulses that give rise to thought and feeling.” (From above article)https://corbettreport.com/the-claude-delusion/