GPT-3 vs. 3M free-text trivia responses • Hacker News Board

Re: GPT-3 vs. 3M free-text trivia responses

#21

Post by hakuseki » Sat, Mar 13, 2021, 3:11 AM UTC

> The more important takeaway: dozens of technologists at IBM spent more than three years and untold millions of dollars building the program specifically trained for Jeopardy! prowess. Less than 10 years later, a general-purpose open-sourced technology without the massive mainframe or cooling fans can compete on the same level.

This statement from the article confused me. GPT-3 is general-purpose, but not open source, nor does it run without massive hardware, nor was it developed without spending millions of dollars.

Re: GPT-3 vs. 3M free-text trivia responses

#22

Post by sprobertson » Sat, Mar 13, 2021, 4:49 AM UTC

bjourne wrote: post #13

fpgaminer wrote: post #10

Somewhat related, but I can't stop thinking about what's going to happen when GPT-4 reads articles like this. Articles like these that discuss GPT-3's output, or comments and articles directly generated by GPT-3, are going to be fed into GPT-4 as part of a theoretical WebText3 dataset. Will that help or hurt GPT-4's learning? What effect will it have? Theoretically you can feed GPT-3's responses back into itself and…

All text generated by current language models is defective. For reasons that are not fully understood, text language models emit is less interesting than text emitted by humans. Unlike humans, they can not score well on both coherence, as in grammatically correct sentences, and interest, as in "having something to say". They are however good at "coherent rambling". Since most language models work by using its outputs…

> For reasons that are not fully understood, text language models emit is less interesting than text emitted by humans.

What's not understood? As you describe, every standard language model is statistically predicting the next word given the previous words, ad infinitum. GPT3 has more parameters to build a more precise "thought" vector, but the goal is still only to produce the next word.

Re: GPT-3 vs. 3M free-text trivia responses

#23

Post by mannykannot » Sat, Mar 13, 2021, 4:54 AM UTC

bjourne wrote: post #13

fpgaminer wrote: post #10

Somewhat related, but I can't stop thinking about what's going to happen when GPT-4 reads articles like this. Articles like these that discuss GPT-3's output, or comments and articles directly generated by GPT-3, are going to be fed into GPT-4 as part of a theoretical WebText3 dataset. Will that help or hurt GPT-4's learning? What effect will it have? Theoretically you can feed GPT-3's responses back into itself and…

All text generated by current language models is defective. For reasons that are not fully understood, text language models emit is less interesting than text emitted by humans. Unlike humans, they can not score well on both coherence, as in grammatically correct sentences, and interest, as in "having something to say". They are however good at "coherent rambling". Since most language models work by using its outputs…

That paper includes this observation:

Why is human-written text not the most probable text? We conjecture that this is an intrinsic property of human language... Grice’s Maxims of Communication (Grice,1975) show that people optimize against stating the obvious.

I don't think that should be at all surprising, if one purpose of speech is to convey information: predictability is the converse of being informational.

Furthermore, to be informative, it is not sufficient for a passage to avoid predictability; it also needs to be pertinent.

Re: GPT-3 vs. 3M free-text trivia responses

#24

Post by blueblisters » Sat, Mar 13, 2021, 5:52 AM UTC

Could GPT-3 lower barriers of entry for search engines? Assuming, it could update its "index" (weights) at regular intervals, it's not too far-fetched to see it competing with Google with high-quality relevant answers to queries.

Re: GPT-3 vs. 3M free-text trivia responses

#25

Post by ronsor » Sat, Mar 13, 2021, 6:00 AM UTC

blueblisters wrote: post #24

Could GPT-3 lower barriers of entry for search engines? Assuming, it could update its "index" (weights) at regular intervals, it's not too far-fetched to see it competing with Google with high-quality relevant answers to queries.

I've tried asking GPT-3 factual questions and sometimes it does great. Other times, well... let's just say it doesn't quite understand. Bad answers are in italics.

Q: How tall is the Eiffel Tower? A: About twenty feet.

Q: What color are the clouds? A: They're a light blue.

Q: Which way does the sun rise in Paris? A: You'd have to be standing on top of a building to see it because it doesn't rise that high.

Q: Who invented the telephone? A: Alexander Graham Bell.

Q: Who invented Google? A: Larry Page and Sergey Brin.

Q: What's the capital of California? A: It's not located in California. It's Sacramento.

Re: GPT-3 vs. 3M free-text trivia responses

#26

Post by ctl0 » Sat, Mar 13, 2021, 6:14 AM UTC

blueblisters wrote: post #24

Could GPT-3 lower barriers of entry for search engines? Assuming, it could update its "index" (weights) at regular intervals, it's not too far-fetched to see it competing with Google with high-quality relevant answers to queries.

There are some works resonate with your idea, here's the first one I can recall.

https://arxiv.org/abs/2012.14983

Re: GPT-3 vs. 3M free-text trivia responses

#27

Post by skybrian » Sat, Mar 13, 2021, 7:54 AM UTC

Der_Einzige wrote: post #16

bjourne wrote: post #13

Earlier quoted context omitted.

All text generated by current language models is defective. For reasons that are not fully understood, text language models emit is less interesting than text emitted by humans. Unlike humans, they can not score well on both coherence, as in grammatically correct sentences, and interest, as in "having something to say". They are however good at "coherent rambling". Since most language models work by using its outputs…

The solution proposed in that paper, nucleolus sampling, effectively solves this problem though...

It’s an odd hack, though.

Adding a bit of randomness results in a more interesting exploration of Borge’s Library of Babel, but it seems weird to assign meaning to the word choice?

Re: GPT-3 vs. 3M free-text trivia responses

#28

Post by knuthsat » Sat, Mar 13, 2021, 8:10 AM UTC

fpgaminer wrote: post #10

Somewhat related, but I can't stop thinking about what's going to happen when GPT-4 reads articles like this. Articles like these that discuss GPT-3's output, or comments and articles directly generated by GPT-3, are going to be fed into GPT-4 as part of a theoretical WebText3 dataset. Will that help or hurt GPT-4's learning? What effect will it have? Theoretically you can feed GPT-3's responses back into itself and…

The whole point of a model is to take advantage of data with more information. If GPT-3 output sucks for predicting human output then it will be ignored by the model during descent. Otherwise the learning algorithm is defective.

Re: GPT-3 vs. 3M free-text trivia responses

#29

Post by notahacker » Sat, Mar 13, 2021, 9:58 AM UTC

fpgaminer wrote: post #10

Somewhat related, but I can't stop thinking about what's going to happen when GPT-4 reads articles like this. Articles like these that discuss GPT-3's output, or comments and articles directly generated by GPT-3, are going to be fed into GPT-4 as part of a theoretical WebText3 dataset. Will that help or hurt GPT-4's learning? What effect will it have? Theoretically you can feed GPT-3's responses back into itself and…

The first part of your statement feels like it's anthropomorphising GPT-4. GPT-4s response to learning that GPT-3 is bad at word play will to store the concept of GPT-3 being bad at word play to recall if prompted about GPT-3 in future, not to swot up on word play. If you want GPT-4 to be better at word play you'll need a corpus of puns, rhymes and higher level concept association and a lot of training on replicating them and when to recognise a prompt might require them.

Re: GPT-3 vs. 3M free-text trivia responses

#30

Post by tsimionescu » Sat, Mar 13, 2021, 10:00 AM UTC

fpgaminer wrote: post #10

Somewhat related, but I can't stop thinking about what's going to happen when GPT-4 reads articles like this. Articles like these that discuss GPT-3's output, or comments and articles directly generated by GPT-3, are going to be fed into GPT-4 as part of a theoretical WebText3 dataset. Will that help or hurt GPT-4's learning? What effect will it have? Theoretically you can feed GPT-3's responses back into itself and…

Your ascribing way too much intelligence to a very simplistic text prediction algorithm. Neither GPT-3 nor GPT-4 (assuming it will just be a bigger GPT-3) have any concept of identity that could influence their predictions in the way you seem to assume.

If GPT-4 would be trained on text produced by GPT-3 and on text that analyzes that, what GPT-4 would learn is how to sound like GPT-3 and like detractors or admirers of GPT-3.

While impressive achievements, GPT-2/3/4/X are just models that predict what the most likely next word is based on a huge corpus of text. There is no model of the world in there, no objects, no identities. Just probabilities of glyphs in certain orders, without any meaning attached.