4 ms·

LLMs seem to be conversational or interactive in non-training phases because they can generate flows of text that mimic conversations, complete with personas, c

LLMs seem to be conversational or interactive in non-training phases because they can generate flows of text that mimic conversations, complete with personas, characters, and personalities. However, this apparent conversation is merely a generative artifact created by the user's input. LLMs do not truly learn or understand; instead, they use their internal data repository to produce "hallucinations" of text that seem conversational. While LLMs do not truly comprehend or understand, their "hallucinations" can be remarkably useful and even enable them to accomplish complex tasks, such as solving mazes or translating languages.

HeatrayEnjoyer 2y ago
They very clearly do have at least some level of understanding. Emergence, and all that.
mewpmewp2 2y ago
What does "truly" mean and how does human understanding differ to make it "truly"?
- outofpaper 2y ago
  Truly means real. Assuming you are a human, you can truly relate and connect ideas with your experiences. You can know what a rose is both in relation to having seen and smelled it and if now having experienced flowers and rose like things. Without qualia you can not know and understand a rose. An LLM no mater how big can never have experienced any of this. The closest it can ever come is to have been trained on the tokens in texts relating to roses. If its more than an LLM it might have also been shown images or even video or roses but no LLM alone has ever seen a rose. LLMs can only contain understanding the way a book can. There is no sentience in a book. Even if we jump to the old analogy of the Chinese room the instruction book in not conscious, nor is the reader following the instructions able to actually speak Chinese. It's beyond and between where lies the entity thar can converse in Chinese.
  - mewpmewp2 2y ago
    So it's about a lack of visual or smell input? Because GPT-4o already can tell what is on images, so it is able to recognize a rose, and of course it has seen many images of roses. It doesn't have the data for the odor, but it seems like arbitrary thing to train it with, for it to have a representation of the odor. Would you consider GPT-4o to truly understand what a rose? And what if it was also trained on odor data?
Sharlin 2y ago
That's using many words to say very little. It reads like a copypasta. Honestly, anyone using the words "truly understand" (or "truly" *anything, really) in any AI conversation should be automatically banned from said discourse. That word combination should be tabooed[1] because it's semantically empty and impossible to define without ending up going in circles. In any case, I fail to see the relevance to my post. I only a) said that NNs can learn categories (an obvious fact) and b) conjectured that the categories "factual" and "creative" may be learnable. [1] https://www.lesswrong.com/posts/WBdvyyHLdxZSAMmoz/taboo-your-words https://www.lesswrong.com/posts/WBdvyyHLdxZSAMmoz/taboo-your...
- amelius 2y ago
  > anyone using the words "truly understand" (or "truly" *anything, really) in any AI conversation should be automatically banned from said discourse What if a psychologist said the same thing about a human? Should they be banned too?
  - outofpaper 2y ago
    Sharlin's response feels knee jerk. And they likely don't truly understand that when conversing with an LLM they are not talking to an LLM but participating in a generative text exchange. LLMs can't ever understand anything anymore than a book can as they are static things. Until we move beyond transformers we're only ever collaborating in the hallucinations.

HeatrayEnjoyer 2y ago

They very clearly do have at least some level of understanding. Emergence, and all that.

mewpmewp2 2y ago

What does "truly" mean and how does human understanding differ to make it "truly"?

outofpaper 2y ago
Truly means real. Assuming you are a human, you can truly relate and connect ideas with your experiences. You can know what a rose is both in relation to having seen and smelled it and if now having experienced flowers and rose like things. Without qualia you can not know and understand a rose. An LLM no mater how big can never have experienced any of this. The closest it can ever come is to have been trained on the tokens in texts relating to roses. If its more than an LLM it might have also been shown images or even video or roses but no LLM alone has ever seen a rose. LLMs can only contain understanding the way a book can. There is no sentience in a book. Even if we jump to the old analogy of the Chinese room the instruction book in not conscious, nor is the reader following the instructions able to actually speak Chinese. It's beyond and between where lies the entity thar can converse in Chinese.
- mewpmewp2 2y ago
  So it's about a lack of visual or smell input? Because GPT-4o already can tell what is on images, so it is able to recognize a rose, and of course it has seen many images of roses. It doesn't have the data for the odor, but it seems like arbitrary thing to train it with, for it to have a representation of the odor. Would you consider GPT-4o to truly understand what a rose? And what if it was also trained on odor data?

Sharlin 2y ago

That's using many words to say very little. It reads like a copypasta. Honestly, anyone using the words "truly understand" (or "truly" *anything, really) in any AI conversation should be automatically banned from said discourse. That word combination should be tabooed[1] because it's semantically empty and impossible to define without ending up going in circles. In any case, I fail to see the relevance to my post. I only a) said that NNs can learn categories (an obvious fact) and b) conjectured that the categories "factual" and "creative" may be learnable. [1] https://www.lesswrong.com/posts/WBdvyyHLdxZSAMmoz/taboo-your-words https://www.lesswrong.com/posts/WBdvyyHLdxZSAMmoz/taboo-your...

amelius 2y ago
> anyone using the words "truly understand" (or "truly" *anything, really) in any AI conversation should be automatically banned from said discourse What if a psychologist said the same thing about a human? Should they be banned too?
- outofpaper 2y ago
  Sharlin's response feels knee jerk. And they likely don't truly understand that when conversing with an LLM they are not talking to an LLM but participating in a generative text exchange. LLMs can't ever understand anything anymore than a book can as they are static things. Until we move beyond transformers we're only ever collaborating in the hallucinations.