Can AI be both RIGHT and WRONG at the SAME TIME? Can we TRUST it?

[AviatorTom]

As a test, a guy asked ChatGPT a simple question, and the AI came back with two contradictory answers.

Question: What song did the HAL 9000 sing in the movie 2001 A Space Odyssey?

AI's Answer: HAL 9000 did not sing a song in the movie "2001: A Space Odyssey." However, HAL 9000 is known for singing the song "Daisy Bell (Bicycle Built for Two)" in the movie "2001: A Space Odyssey."

The Truth: The singing occurs during a dramatic scene where HAL is being deactivated by astronaut Dave Bowman.

What do you think? Should we trust AI if it's able to simultaneously give contradictory information?

[TonyMetal___86]

Don't expect something perfect created from an unperfect being which is us...
If one day the A. I reaches an advanced level such like in the movies of terminator and the matrix than we might be doomed or used as slaves!

Since our whole current technology is based on sci-fi and brought from star trek, than you can get a glimpse of how it may look like in the future, for example look at DATA from star trek the next generation, his skills far more exceeds the humans skills and he was on a higher level in nearly all the different aspects, still he always wish to have feelings like humans until they planted a chip inside of him to make him feel like humans, also how many times did he malfunctioned and how many times he turned up against his makers?

Now that's a movie and the script writer decides if it ends good or bad with the agreement of the director but in real life, you can't expect a happy ending if an A. I went wrong and you couldn't fix the issue, especially if your lethal weapons are under it's control, which means a simple malfunction might lead to an appocalypse...

[TonyMetal___86]

Tnx for the mho

[Craftsman]

My opinion is that all AI needs to be checked by humans. Too many such errors are coming to light these days. It seems if AI cannot find data, it starts fabricating, and that can be seen as lying, fibbing, and pulling facts from the air (or elsewhere, though AI does not have such an anatomical orifice.)

[IslaTheWitch]

ChatGPT can make mistakes. Consider checking important information.

It's written on every service which is based on chatGPT. Why? Because so called AI still learns. There is a huge difference in answers from the old free v3.5 and paid one v4.

I took the liberty to ask your question to the chatGPT v4 chatbot. Here is the answer:

In the movie "2001: A Space Odyssey," the HAL 9000 computer sings the song "Daisy Bell" (also known as "Bicycle Built for Two"). This occurs during a scene where HAL's cognitive functions are being gradually shut down. The choice of this song is particularly significant because "Daisy Bell" was the first piece of music to be sung by a computer using speech synthesis. This occurred in 1961 at Bell Labs, using an IBM 704 computer, which makes the inclusion of the song in the film a nod to this milestone in computer history.

[LadyKitty]

I see AI as a tool but if you are looking for answers i'd suggest to double check what it considers factually correct.

[Nikki1989]

I can spot AI generated text from a mile away. Humans don't naturally write like that. Whenever I see it on here, I report it as spam.

[OlderAndWiser]

I never suggested that AI should be trusted. We are moving forward far too fast!

[HouseMoat]

AI can be both good and bad

The problems going to become, how good AI is going to get, because it's going to become VERY hard to distinguish what is real and what is fake news

[AngryCarl]

AI doesn't know the difference between right and wrong. You can tell AI that black is white and it will react accordingly.

[Guido]

We cannot trust humans and they do the programming so it makes sense not to trust AI.

[AnusNdaPainus]

Even google makes mistakes