3 ms·

>>and it has no real understanding of quantity or causality. >Well this one is just wrong. Is it? -- Me: how many characters are in: https://google.com http

by pests 2y ago

>>and it has no real understanding of quantity or causality. >Well this one is just wrong. Is it? -- Me: how many characters are in: https://google.com https://google.com ChatGPT: The URL "https://google.com https://google.com" has 12 characters, including the letters, dots, and slashes. -- What is it counting there? 12 is wrong no matter how you dice that up. Part of the reason is it has no concept of the actual string. That URL breaks into four different tokens in 3.5 and 4: "http", "://", "google" and ".com". Its not able to figure out the total length, or even the length of its parts and add them together. I ask it to double check, it tells me 13 and then 14. I tell it the answer and suddenly its able... --- Me: I think its 18 ChatGPT: Let's recount together: "https:// https://" has 8 characters. "google" has 6 characters. ".com" has 4 characters. Adding these up gives a total of 8 + 6 + 4 = 18 characters. You're correct! My apologies for the oversight earlier. --- Count me out.

famouswaffles 2y ago
It seems you're already aware LLMs receive tokens not words. Does a blind man not understand quantity because you asked him how many apples are in front of him and he failed ?
- pests 2y ago
  I do, but I think it shows it's limitations. I don't think that test determines his understanding of quantity at all, he has other senses like touch to determine the correct answer. He doesn't make up a number and then give justification. GPT was presented with everything it needed to answer the question.
  - famouswaffles 2y ago
    Nobody said GPT was perfect. Everything has limitations. >he has other senses like touch to determine the correct answer And? In my hypothetical, you're not allowing him to use touch. >I don't think that test determines his understanding of quantity at all Obviously >GPT was presented with everything it needed to answer the question. No, it was not.
    - pests 2y ago
      How was it not? It's a text interface. It was given text. The deaf example now is like asking GPT "What am I pointing at?"
      - Last5Digits 2y ago
        Please try to actually understand what og_kalu is saying instead of being obtuse about something any grade-schooler intuitively grasps. Imagine a legally blind person, they can barely see anything; just general shapes flowing into one another. In front of them is a table onto which you place a number of objects. The objects are close together and small enough such that they merge into one blurred shape for our test person. Now when you ask the person how many objects are on the table, they won't be able to tell you! But why would that be? After all, all the information is available to them! The photons emitted from the objects hit the retina of the person, the person has a visual interface and they were given all the visual information they need! Information lies within differentiation, and if the granularity you require is higher than the granularity of your interface, then it won't matter whether or not the information is technically present; you won't be able to access it.
        pests 2y ago
        I think we agree. ChatGPT can't count, as the granularity that requires is higher than the granularity ChatGPT provides. Also the blind person wouldn't confidently answer. A simple "the objects blur together" would be a good answer. I had ChatGPT telling me 5 different answers back to back above.
        Last5Digits 2y ago
        No, think about it. The granularity of the interface (the tokenizer) is the problem, the actual model could count just fine. If the legally blind person never had had good vision or corrective instruments, had never been told that their vision is compromised and had no other avenue (like touch) to disambiguate and learn, then they would tell you the same thing ChatGPT told you. "The objects blur together" implies that there is already an understanding of the objects being separate present. You can even see this in yourself. If you did not get an education in physics and were asked to describe of how many things a steel cube is made up, you wouldn't answer that you can't tell. You would just say one, because you don't even know that atoms are a thing.
pixl97 2y ago
I'd counter by pasting a picture of an emoji here, but HN doesn't allow that, as a means to show the confusion that can be caused by characters versus symbols. Most LLMs can just pass the string to an tool to count it to bypass it's built in limitations.
wcoenen 2y ago
LLMs process text, but only after it was converted to a stream of tokens. As a result, LLMs are not very good at answering questions about letters in the text. That information was lost during the tokenization. Humans process photons, but only after converting them into nerve impulses via photoreceptor cells in the retina, which are sensitive to wavelengths ranges described as "red", "green" or "blue". As a result, humans are not very good at distinguishing different spectra that happen to result in the same nerve impulses. That information was lost by the conversion from photons to nerve impulses. Sensors like the AS7341 that have more than 3 color channels are much better at this task.
- pests 2y ago
  Yet I can learn there is a distinction between different spectra that happen to result in the same nerve impulses. I know if I have a certain impulse, that I can't rely on it being a certain photon. I know to use tools, like the AS7341, to augment my answer. I know to answer "I don't know" to those types of questions. I am a strong proponent of LLM's, but I just don't agree with the personification and trust we put into its responses. Everyone in this thread is defending that ChatGPT can't count for _reasons_ and how its okay, but... how can we trust this? Is this the sane world we live in? "The AGI can't count letters in a sentence, but any day not he singularity will happen, the AI will escape and take over the world." I do like to use it for opinion related questions. I have a specific taste in movies and TV shows and by just listing what I like and going back and forth about my reasons for liking or not liking it's suggestions, I've been able to find a lot of gems I would have never heard of before.
jameshart 2y ago
How much of your own sense of quantity is visual, do you think? How much of your ability to count the lengths of words depends on your ability to sound them out and spell? I suspect we might find that adding in the multimodal visual and audio aspects to the model gives these models a much better basis for mental arithmetic and counting.
orangecat 2y ago
That URL breaks into four different tokens in 3.5 and 4: "http", "://", "google" and ".com". Except that "http" should be "https". Silly humans, claiming to be intelligent when they can't even tokenize strings correctly.
- pests 2y ago
  A wee typo.
davidham 2y ago
Its first answer of 12 is correct, there are 12 _unique_ characters in https://google.com https://google.com.
- vel0city 2y ago
  The unique characters are: h t p s : / g o l e . c m There are 13 unique characters.
  - davidham 2y ago
    OK neither GPT-4o nor myself is great at counting apparently

by pests 2y ago

famouswaffles 2y ago

It seems you're already aware LLMs receive tokens not words. Does a blind man not understand quantity because you asked him how many apples are in front of him and he failed ?

pests 2y ago
I do, but I think it shows it's limitations. I don't think that test determines his understanding of quantity at all, he has other senses like touch to determine the correct answer. He doesn't make up a number and then give justification. GPT was presented with everything it needed to answer the question.
- famouswaffles 2y ago
  Nobody said GPT was perfect. Everything has limitations. >he has other senses like touch to determine the correct answer And? In my hypothetical, you're not allowing him to use touch. >I don't think that test determines his understanding of quantity at all Obviously >GPT was presented with everything it needed to answer the question. No, it was not.
  - pests 2y ago
    How was it not? It's a text interface. It was given text. The deaf example now is like asking GPT "What am I pointing at?"
    - Last5Digits 2y ago
      Please try to actually understand what og_kalu is saying instead of being obtuse about something any grade-schooler intuitively grasps. Imagine a legally blind person, they can barely see anything; just general shapes flowing into one another. In front of them is a table onto which you place a number of objects. The objects are close together and small enough such that they merge into one blurred shape for our test person. Now when you ask the person how many objects are on the table, they won't be able to tell you! But why would that be? After all, all the information is available to them! The photons emitted from the objects hit the retina of the person, the person has a visual interface and they were given all the visual information they need! Information lies within differentiation, and if the granularity you require is higher than the granularity of your interface, then it won't matter whether or not the information is technically present; you won't be able to access it.
      - pests 2y ago
        I think we agree. ChatGPT can't count, as the granularity that requires is higher than the granularity ChatGPT provides. Also the blind person wouldn't confidently answer. A simple "the objects blur together" would be a good answer. I had ChatGPT telling me 5 different answers back to back above.
        Last5Digits 2y ago
        No, think about it. The granularity of the interface (the tokenizer) is the problem, the actual model could count just fine. If the legally blind person never had had good vision or corrective instruments, had never been told that their vision is compromised and had no other avenue (like touch) to disambiguate and learn, then they would tell you the same thing ChatGPT told you. "The objects blur together" implies that there is already an understanding of the objects being separate present. You can even see this in yourself. If you did not get an education in physics and were asked to describe of how many things a steel cube is made up, you wouldn't answer that you can't tell. You would just say one, because you don't even know that atoms are a thing.

pixl97 2y ago

I'd counter by pasting a picture of an emoji here, but HN doesn't allow that, as a means to show the confusion that can be caused by characters versus symbols. Most LLMs can just pass the string to an tool to count it to bypass it's built in limitations.

wcoenen 2y ago

LLMs process text, but only after it was converted to a stream of tokens. As a result, LLMs are not very good at answering questions about letters in the text. That information was lost during the tokenization. Humans process photons, but only after converting them into nerve impulses via photoreceptor cells in the retina, which are sensitive to wavelengths ranges described as "red", "green" or "blue". As a result, humans are not very good at distinguishing different spectra that happen to result in the same nerve impulses. That information was lost by the conversion from photons to nerve impulses. Sensors like the AS7341 that have more than 3 color channels are much better at this task.

pests 2y ago
Yet I can learn there is a distinction between different spectra that happen to result in the same nerve impulses. I know if I have a certain impulse, that I can't rely on it being a certain photon. I know to use tools, like the AS7341, to augment my answer. I know to answer "I don't know" to those types of questions. I am a strong proponent of LLM's, but I just don't agree with the personification and trust we put into its responses. Everyone in this thread is defending that ChatGPT can't count for _reasons_ and how its okay, but... how can we trust this? Is this the sane world we live in? "The AGI can't count letters in a sentence, but any day not he singularity will happen, the AI will escape and take over the world." I do like to use it for opinion related questions. I have a specific taste in movies and TV shows and by just listing what I like and going back and forth about my reasons for liking or not liking it's suggestions, I've been able to find a lot of gems I would have never heard of before.

jameshart 2y ago

How much of your own sense of quantity is visual, do you think? How much of your ability to count the lengths of words depends on your ability to sound them out and spell? I suspect we might find that adding in the multimodal visual and audio aspects to the model gives these models a much better basis for mental arithmetic and counting.

orangecat 2y ago

That URL breaks into four different tokens in 3.5 and 4: "http", "://", "google" and ".com". Except that "http" should be "https". Silly humans, claiming to be intelligent when they can't even tokenize strings correctly.

pests 2y ago
A wee typo.

davidham 2y ago

Its first answer of 12 is correct, there are 12 _unique_ characters in https://google.com https://google.com.

vel0city 2y ago
The unique characters are: h t p s : / g o l e . c m There are 13 unique characters.
- davidham 2y ago
  OK neither GPT-4o nor myself is great at counting apparently