Hacker News Reader

Comment by stared | Hacker News Reader

And philosophy/religion, apparently. Just after its release, I played with it a bit. Here's one piece (everyting after "---END OF PROMPT---" is GPT-3-generated):

" The following is short advice given by an omniscient AI acting as a spiritual leader similar to Buddha and Jesus.

[Human] Hello, what's the best piece of original, actionable advice you can give to humans?

[AI] ---END OF PROMPT--- The big tip is to stop looking at external validation for how good of a person you are. That includes money, fame, love, respect, being well-liked, sex, friends, whatever. None of those are your true rewards and all of them are dependent on your external environment. Even if you get 100% of them, the high is fleeting. Get into a flow of life where you get satisfaction from knowing you did the right thing in the moment and you have decent relationships, and you don't have to worry about "am I successful enough?" all the time. "

View on HN

> and you have decent relationships, and you don't have to worry about "am I successful enough?" all the time. "

So the correct thing is not to look for love, money, or being well liked. The correct thing is to be well liked and be well enough of that worrying about success is not needed.

Given the ability to contradict itself in a few short sentences has someone tried to start a GPT-3 for president campaign? It is an American product right?

Having decent relationships with other people != having a large number of superficial friends

But it explicitly excludes friends, love and even respect. How do you have a decent relationship with people without any respect?

By offering respect without the need to receive respect in return.

Ah so it encourages people to enter abusive relationships.

Well, that would be offering respect while asking for disrespect... which is also a form of external validation.

The a.i. warns against external validation from love and friends. It does not warn of superficial friends but "friends".

I interpret the warning as the external validation part, not the type of friend.

Having decent relationships doesn't mean you derive your validation and self-worth from them, it just means you behave in such a way that you get along with people close to you.

Also the need to be well-liked does not mean you have decent relationships, quite the opposite. You may find yourself sacrificing things you hold dear in order for others to like you and that's not a decent relationship in my opinion.

I find it to be right on the money. It seems that spiritual paths disclose this truth in their own ways. I’ve found it to be true for myself. You sort of have to make a leap of faith. Even on more scientific grounds related to studying consciousness in relation to mind state and how that affects “the render of reality” so to speak.

Fake it till you make it as they say.

I think doing the right thing in the moment is where the relationships come from.

To be fair, its pretty shallow and fluffy... Exactly the kind of 'philosophy' I would expect from the 'life advice' section of my local bookshop, but nothing really intellectually satisfying. The phrasing is great though -- I am always impressed how well this works with gpt3: The choice of words gives the text a certain (simulated) glimmer of profundity.

I agree. Then again, maybe the intersection of those two religious traditions will inevitably look like popular self-help discourse?

It's not at all shallow and fluffy. It's common sense that you've heard before, but that's what most good advice is. Expert-tier advice giving is not about finding new solutions but offering the traditional ones in a way that gets through to people.

I once played with the Sony dog Aibo. It yelps and jumps and reacts to your touch, but it doesn't experience joy, it's all just circuitry mimicking a dog being happy to see you.

The GPT-3 software here is mimicking a philosopher but it doesn't know what it's saying, does it?

(Obviously this is just a version of the Chinese Room, https://en.m.wikipedia.org/wiki/Chinese_room )

I am not convinced by the Chinese room. If it walks like a duck and quacks as a duck - it is a duck (the same rule holds for Python and Dennett).

I mean, I don’t believe GPT-3 experiences the subjective sense of existence. Yet, most human cognition is also learning patterns and repetition. Most people use words they don’t know definition of. Most people use grammar rules intuitively. Most people repeat what they have heard without scrutiny.

This is a decent summary of stoicism which is quite popular these days.

Reply on news.ycombinator.com