Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Ask HN: How does ChatGPT work?

news.ycombinator.com

Post reply on HN
Print view View original

71–80 of 185 posts

Re: Ask HN: How does ChatGPT work?

#71

Post by zozbot234 »

undefuser wrote: post #69
DemocracyFTW2 wrote: post #62

What baffles and makes it hard for me to wrap my brain around the statement that ChatGPT is "just predictive text analysis" is that it certainly feels like it goes into detail, picks up and repeats the particulars, presents a line of reasoning that sounds, well, reasonable, until you realize it got a number of things terribly wrong, often in deceptively subtle ways: Me: Which number is bigger, 873593742593454623 or 8…

I applied the same prompts and the results get better: Me: Which number is bigger, 873593742593454623 or 873593742593554623 ChatGPT: These two numbers are equal. The first number is simply the second number with an extra digit added at the end. Since the added digit does not affect the value of the number, both numbers are equal. Me: Apply your reasoning and finding the bigger of two numbers to compare 123 and 53. Ch…

ChatGPT is a master troll.

Re: Ask HN: How does ChatGPT work?

#73

Post by bheadmaster »

Scarblac wrote: post #65
int_19h wrote: post #64

Earlier quoted context omitted.

I don't know what it uses to output non-English text, but if you address it in a different language, it'll quite happily speak that back to you. And I'd say it understands it just fine, not really any worse than English - although it sometimes gets the word forms wrong in inflective languages.

But I was asking it for rhyming words in Dutch (because of Sinterklaas), and it suggested words that would rhyme if pronounced as English...

I also found the same thing happening in my native tongue. It would be much more exciting if it had "native" understanding of foreign languages.

Re: Ask HN: How does ChatGPT work?

#74

Post by tluyben2 »

mclightning wrote: post #41
akelly wrote: post #18

The way they went from GPT-3 to ChatGPT is really quite genius. My understanding is that it's something like this: 1. Start with GPT-3, which predicts the next word in some text and is trained on all the text on the internet 2. Take thousands of prompts, generate several responses for each of them, and have human reviewers rank the responses for each prompt from best to worst 3. The GPT model needs a massive amount o…

That's super interesting. When GPT-3 came out, I wrote an article inspired by it. That we could one day build an AI that acts like AGI, by a crazy vast amount of multimedia training data, collected by willing users to participate in ever improving AI interactions; https://medium.com/swlh/bicameral-mind-humanoid-robot-with-g...

When I was a first year AI student beginning of the 90s I asked my professor what would happen if we just made a massive neural network and trained it with all information in the world. He said it cannot happen as it it impossible.

Re: Ask HN: How does ChatGPT work?

#75

Post by visarga »

akelly wrote: post #18

The way they went from GPT-3 to ChatGPT is really quite genius. My understanding is that it's something like this: 1. Start with GPT-3, which predicts the next word in some text and is trained on all the text on the internet 2. Take thousands of prompts, generate several responses for each of them, and have human reviewers rank the responses for each prompt from best to worst 3. The GPT model needs a massive amount o…

> 2. Take thousands of prompts, generate several responses for each of them, and have human reviewers rank the responses for each prompt from best to worst

Step 2 is not that. It's manually writing responses for a few tasks.

> A labeller demonstrates the desired output behavior.

(left side on https://cdn.openai.com/chatgpt/draft-20221129c/ChatGPT_Diagr...)

So it is supervised training in this stage. Ranking is the next stage, for training the reward model. This is not the reward model, it's a model to generate sample responses to be used by the reward model.

So there are two kinds of manual work involved here - manually demonstrating how to solve tasks, and ranking responses. There is even talk about how much effort to invest in the first vs the second and what is the trade-off.

Re: Ask HN: How does ChatGPT work?

#76

Post by tluyben2 »

starik36 wrote: post #38
akelly wrote: post #18

The way they went from GPT-3 to ChatGPT is really quite genius. My understanding is that it's something like this: 1. Start with GPT-3, which predicts the next word in some text and is trained on all the text on the internet 2. Take thousands of prompts, generate several responses for each of them, and have human reviewers rank the responses for each prompt from best to worst 3. The GPT model needs a massive amount o…

Ok, but how does it take my code and fixes the bug? It's my own code, no one has seen it besides me and the model wasn't trained on it.

It often does[0], I am doing more experiments currently by writing the fixed code back in place and running tests automatically.

[0] https://brainfisheatfishbrain.com/post/chatgpt-code-reviews/

Re: Ask HN: How does ChatGPT work?

#78

Post by namaria »

tluyben2 wrote: post #74
mclightning wrote: post #41

Earlier quoted context omitted.

That's super interesting. When GPT-3 came out, I wrote an article inspired by it. That we could one day build an AI that acts like AGI, by a crazy vast amount of multimedia training data, collected by willing users to participate in ever improving AI interactions; https://medium.com/swlh/bicameral-mind-humanoid-robot-with-g...

When I was a first year AI student beginning of the 90s I asked my professor what would happen if we just made a massive neural network and trained it with all information in the world. He said it cannot happen as it it impossible.

This reminds me of a major US newspaper declaring heavier-than-air flying machines a million years away mere months before the Wright brothers experiments.

Re: Ask HN: How does ChatGPT work?

#79

Post by karmasimida »

timonoko wrote: post #42

How does the non-english languages part work? I thought maybe they use Google translator, but remembered that Russians have trained it to not to understand "russophobic" sentences. -- Mitä tarkoittaa ryssänvastainen, explain in English. -- Ryssänvastainen means "anti-Russian" or "anti-Russian sentiment." It refers to an attitude or behavior that is hostile or opposed to Russia or Russian interests.

Just vast amount of webpages in other languages.

Re: Ask HN: How does ChatGPT work?

#80

Post by tnzk »

akelly wrote: post #18

The way they went from GPT-3 to ChatGPT is really quite genius. My understanding is that it's something like this: 1. Start with GPT-3, which predicts the next word in some text and is trained on all the text on the internet 2. Take thousands of prompts, generate several responses for each of them, and have human reviewers rank the responses for each prompt from best to worst 3. The GPT model needs a massive amount o…

> 6. Release ChatGPT to the public, and use user feedback like response upvotes/downvotes to further optimize the reward model, while continuing to train ChatGPT against the reward model

Can someone provide a pointer to an article that elaborate this part?

Post reply on HN