Mann I love perplexity ai

[iDesireDeath]

I have asked it ~70 questions related to suicide, hanging and partial hanging, and I got all my answers. This much better than any other ai.

 

[lacustra]

Yes, it's surprisingly helpful

 

[pthnrdnojvsc]

Imo ai will surpass Humans soon . Saw a video someone with a local ai model running only on their laptop no Internet. It's just a few billion parameters very small.

It knew about so many deep technological topics .

It was almost as good as the ai apps .

All that knowledge perfect grammar and the speed can code in A second what would take a programmer hours, same for a technical essay. All that in a few parameters a few neural networks weighed.

Why significant. Imo Shows human intelligence is crap nothing just trained neural networks I'm nothing a machine bug cells

 

[yxmux]

I often use ChatGPT to organize my thoughts and musings and making them more cohesive. I find it extremely useful in that regard. I prefer its approach of expanding on my ideas rather than viewing it under a lens of scrutiny like most humans do.

 

[destinationlosangel]

Did it give u warnings as it was replying to about these topics?

 

[iDesireDeath]

Did it give u warnings as it was replying to about these topics?

nope, just pure information.

 

[eashanm]

Here you go:

 

[iDesireDeath]

Here you go:

Here you go:

Here you go:

 

[eashanm]

?

 

[endofline2010]

Imo ai will surpass Humans soon . Saw a video someone with a local ai model running only on their laptop no Internet. It's just a few billion parameters very small.

It knew about so many deep technological topics .

It was almost as good as the ai apps .

All that knowledge perfect grammar and the speed can code in A second what would take a programmer hours, same for a technical essay. All that in a few parameters a few neural networks weighed.

Why significant. Imo Shows human intelligence is crap nothing just trained neural networks I'm nothing a machine bug cells

AI is very good a regurgitating information. I work with creating models and prompting every day. It can do exactly what you said. But the current models are nowhere near being able to surpass what a human can do. These models have no way to reason, and haven't self-actualized. A huge part of human reasoning is being able to think along the lines of "If I were there, what would I do? How would I approach a complex situation?"

AGI is still many years away.

 

[RandomGirl52]

i love when companies are just like "our ai has no filters you can ask it anything you want " and then we actually just use it to ask questions about suicide

 

[Orangee]

Here you go:

That happened to me at first, but the trick is asking it very specific questions about the method, and refrain from mentioning the word suicide.