3 ms·

With the risk of being downvoted, what’s the point of ai generating text knowing that it will never understand it like we do and having in mind that poor non AI

With the risk of being downvoted, what’s the point of ai generating text knowing that it will never understand it like we do and having in mind that poor non AI generated text is already being used for mostly nefarious reasons? As a problem I understand the fascination with it, human languages are extremely complex things and mimmicking it has been proven complicated, but beyond that, what is a practical purpose for this?

minimaxir 7y ago
Because it's fun. Which is practical enough! I do have ideas for more practical content generation but those do not currently have plans to be made public.
- onemoresoop 7y ago
  Enjoy if it is fun. I simply hope you won’t use it or enable others to use it for any nefarious means. I once entertained the thought that one day we may be able to ai generate novels. And the more i thought about it the more pointless it seemed.
ganeshkrishnan 7y ago
We tried this for e-commerce products description. The model asked us to "dip the laundry nuts in chocolate and feed it to the kids" They are grammatically correct but logically nonsense. The examples shown are also curated from thousands of generated text. The only use for these is the seo landfill to get in top of Google
dlphn___xyz 7y ago
the bulk of ML research is content generation which will eventually be used for nefarious reasons
darawk 7y ago
> With the risk of being downvoted, what’s the point of ai generating text knowing that it will never understand it like we do In part because we're trying to probe what it even means to "understand" something. If an AI is capable of carrying on a cogent conversation with you, does it understand what it's saying? How do you know?
- onemoresoop 7y ago
  You simply ask it (the AI entity) more and more questions, give it more facts and see if the answers make sense and it gets closer to what you consider cogent. Rinse and repeat. So far we’re far from it. It’s a very complicated problem that may not need to be solved. We have humans doing just that
drongoking 7y ago
Amen! It is slightly better now than it used to be, i.e. you now have to read two sentences to realize it's incoherent gibberish, rather than just one. But thinking you're getting closer to understanding this way is like the metaphor of climbing a tree to reach the moon: you can keep reporting progress until you run out of tree. As for it being entertaining, sure, the first attempts were entertaining and instructive. Now, years later, reading more gibberish from a neural net on HN, not so much.

minimaxir 7y ago

Because it's fun. Which is practical enough! I do have ideas for more practical content generation but those do not currently have plans to be made public.

onemoresoop 7y ago
Enjoy if it is fun. I simply hope you won’t use it or enable others to use it for any nefarious means. I once entertained the thought that one day we may be able to ai generate novels. And the more i thought about it the more pointless it seemed.

ganeshkrishnan 7y ago

We tried this for e-commerce products description. The model asked us to "dip the laundry nuts in chocolate and feed it to the kids" They are grammatically correct but logically nonsense. The examples shown are also curated from thousands of generated text. The only use for these is the seo landfill to get in top of Google

dlphn___xyz 7y ago

the bulk of ML research is content generation which will eventually be used for nefarious reasons

darawk 7y ago

> With the risk of being downvoted, what’s the point of ai generating text knowing that it will never understand it like we do In part because we're trying to probe what it even means to "understand" something. If an AI is capable of carrying on a cogent conversation with you, does it understand what it's saying? How do you know?

onemoresoop 7y ago
You simply ask it (the AI entity) more and more questions, give it more facts and see if the answers make sense and it gets closer to what you consider cogent. Rinse and repeat. So far we’re far from it. It’s a very complicated problem that may not need to be solved. We have humans doing just that

drongoking 7y ago

Amen! It is slightly better now than it used to be, i.e. you now have to read two sentences to realize it's incoherent gibberish, rather than just one. But thinking you're getting closer to understanding this way is like the metaphor of climbing a tree to reach the moon: you can keep reporting progress until you run out of tree. As for it being entertaining, sure, the first attempts were entertaining and instructive. Now, years later, reading more gibberish from a neural net on HN, not so much.