5 ms·

Its not like ChatGPT made this up. There were pre-existing YouTube tutorials and python scripts available that used OpenCage an purported to do this. OpenCage e

Its not like ChatGPT made this up. There were pre-existing YouTube tutorials and python scripts available that used OpenCage an purported to do this. OpenCage even blogged about this problem almost a year ago[1]. Honestly it looks more like OpenCage is trying to rehash the same issue for more clicks by spinning it off the hugely popular ChatGPT keywords. Wouldn't be too surprised if they created the original python utilities themselves just to get some publicity by denouncing them. 1. https://blog.opencagedata.com/post/we-can-not-convert-a-phone-number-into-a-location-sorry https://blog.opencagedata.com/post/we-can-not-convert-a-phon...

ceejayoz 4y ago
That seems like a pretty nasty assertion to bandy around with zero evidence.
- singlow 4y ago
  I cannot think of any other reason why the new blog post wouldn't have mentioned the obvious connection to the earlier issues that they had. They want to make it seem like ChatGPT invented this use case but they know that the sample code that ChatGTP learned from was mentioned in their previous blog post.
  - ceejayoz 4y ago
    There's a vast chasm between "whoever wrote this article didn't think to link to a similar issue a year ago" and "the first incident was a malicious hoax".
    - singlow 4y ago
      [flagged]
      - ceejayoz 4y ago
        That's another apparently evidence-free accusation. Is there some undisclosed bad blood here?
        singlow 4y ago
        No history at all. Do you have any undisclosed relationship? Hmm. My post was getting traction then suddenly its downvoted to the bottom. Maybe it was a social media management platform. Do you know anyone who runs one of those?
        dang 4y ago
        You broke the site guidelines pretty badly in this thread, even stooping to personal attacks. We ban accounts that do that, so please don't do that. https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html
      - mtmail 4y ago
        Ed is my co-founder, he writes all our blog posts because I suck I writing. He also does more than half of our podcast episodes https://thegeomob.com/podcast https://thegeomob.com/podcast (the guy on the left). Last I saw him (yesterday) he was real.
      - freyfogle 4y ago
        Just re-checked the org chart. There's no social media personal assistant.
    - luckylion 4y ago
      I don't understand the original comment to suggest that. Rather: it's a known issue. ChatGPT does nothing new, and certainly doesn't do it by itself -- it just rehashes what others have already written. Like Google might send you visitors for something that's not even present on your website because others link to you mentioning it. What the comment suggested was that they're now bringing this up again to get attention (and links) since it's combined with ChatGPT. That's not "malicious", but it's also not exactly "wow, we just realized this happens".
      - seszett 4y ago
        What the comment suggested is that the company deliberately created tools using their own API in a wrong way in order to write a blog post about it. If that's not an accusation of being malicious I don't know what could be.
- vlunkr 4y ago
  There's also no clear motive. They want to attract users to a fake feature their free tier?
freyfogle 4y ago
Hi, Ed from OpenCage here, author of the post. We do have python tutorials and SDKs showing how to use our service for ... geocoding, the actual service we provide. I wrote the post mainly to have a page I can point people to when they ask why "it isn't working". Rather than take the user through a tour of past posts I need something simple they will hopefully read. But fair point, I can add a link to last year's post about the erronious youtube tutorials as well. What I think you can't appeciate is the difference of scale. A faulty youtube video drives a few users. In the last weeks ChatGPT is sending us several orders of magnitude more frustrated sign-ups.
- singlow 4y ago
  I get frustrated at the number of things ChatGPT gets blamed for that aren't its fault. It is completely understandable that if there are repos out on GitHub like the one for Phomber[1] thant ChatGPT would find that code and have no idea that it was phoney. Suggesting that ChatGPT just made this up out of thin air when you know it didn't is not very responsible. 1. https://github.com/s41r4j/phomber https://github.com/s41r4j/phomber
  - jraph 4y ago
    You are blaming the victim. OpenAI is to be blamed. They know what they are doing. They provide something that sounds over-confident for anything it says, knowing full well that it can't actually know if what it generated is accurate because it is designed to generate plausible sentences using statistics and probabilities, not verified facts from a database. On top of it, they trained it on an uncontrolled set of texts (though IIUC even a set of verified text would not be enough, nothing guarantees that a LM would produce correct answers). And they provide it to the general population, which doesn't always understand very well how it works and, above all, its limitations. Including developers. Few people actually understand this technology, including myself. Inevitably, it was going to end up causing issues. This post factually presents a problematic situation for the authors of this post. How ChatGPT works or how it can end up producing wrong results is irrelevant to the post's authors problem. It just does, and it causes troubles because of the way OpenAI decided to handle things. And it's not "fair enough, because this false stuff can be found on the internet".
gus_massa 4y ago
That explains why ChatGPT is confused. It may be an old problem, but I guess users are more use to a random YouTube video with wrong information. But the computer is always right so ChatGPT is always right, so users may be more annoyed to discover that the recommendation is wrong and blame them instead of ChatGPT.
fwlr 4y ago
Devs making baby’s first mobile app add “request location information” permissions, the devices start giving them the phone’s GPS information in the form of lat/lon pairs, and those devs naturally look for a service to make that data useful. What they want is “reverse geolocation”, i.e. take a lat/lon pair and return information that makes sense to a human (country, state, nearby street address, etc). This is a service that OpenCage provides, and for whatever reason OpenCage happens to be one of the popular services for this use case. (Maybe it’s because you get the text description of location back right away without having to do a round trip through a heavyweight on-screen map, maybe their free tier allows more requests than most, maybe their api is easier to use, maybe they are lucky or skilled with SEO and their tutorial happens to be the first result for some common phrases, who knows.) So there’s this process that starts with a search for “convert phone location to address”, often involves the OpenCage api, and ends with a happy developer getting the information they wanted. Various algorithms pick up on the existence and repeated traversal of this happy path. In another part of the internet, code tutorial content farms notice a demand for determining an incoming call’s location from the number that’s calling. They search for things like “convert phone number to location” and “convert phone number to address”. Some of these searches end up falling into the nearby well-trodden path of “convert phone location to address” and the content farmer is presented with the OpenCage api. They mess around with the api for a bit and find they can start from a phone number and get a successful api call that returns a lat/lon pair. A successful api call that returns legitimate-looking lat/lon data is all they need to make a video, they make it and post it. Higher-quality, more scrupulous code tutorials attempt to answer this same demand but find it’s not possible, so those tutorials don’t get made, leaving the less scrupulous ones that stop with a successful-looking api call to flourish in this space. The tutorial is doing well, so the content farms endlessly recycle it into blogspam. As a result, OpenCage starts getting weird usage patterns, tracks them down, finds the source is these tutorials, and makes a post about it. Some time later, ChatGPT is released. People are astounded with its ability to write code and start using it for this purpose. Naturally, some of those people have the same demand as the previous generation of devs who stumbled onto the unscrupulous code tutorials. Because of the blogspam, ChatGPT’s training data includes many variations on the tutorial, and just as naturally it ends up reproducing that tutorial when asked - except ChatGPT’s magic kicks in and instead of including (what its embeddings see as) some weird unrelated area-code-to-string nonsense from the tutorial, it just bullshits some plausible-sounding data plumbing code instead. Unfortunately, because the tutorial never worked in the first place, that weird hacky irrelevant bit that ChatGPT ignored happened to be the secret sauce that makes the whole thing superficially appear to work. As a result, OpenCage starts getting weird usage patterns, tracks them down, finds the source is ChatGPT, and makes a post about it. In deference to Hacker News’ policy of keeping comments pleasant, I will elide the analysis of the process that leads to comments accusing OpenCage of nefariously engineering the whole thing for attention.

ceejayoz 4y ago

That seems like a pretty nasty assertion to bandy around with zero evidence.

singlow 4y ago
I cannot think of any other reason why the new blog post wouldn't have mentioned the obvious connection to the earlier issues that they had. They want to make it seem like ChatGPT invented this use case but they know that the sample code that ChatGTP learned from was mentioned in their previous blog post.
- ceejayoz 4y ago
  There's a vast chasm between "whoever wrote this article didn't think to link to a similar issue a year ago" and "the first incident was a malicious hoax".
  - singlow 4y ago
    [flagged]
    - ceejayoz 4y ago
      That's another apparently evidence-free accusation. Is there some undisclosed bad blood here?
      - singlow 4y ago
        No history at all. Do you have any undisclosed relationship? Hmm. My post was getting traction then suddenly its downvoted to the bottom. Maybe it was a social media management platform. Do you know anyone who runs one of those?
        dang 4y ago
        You broke the site guidelines pretty badly in this thread, even stooping to personal attacks. We ban accounts that do that, so please don't do that. https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html
    - mtmail 4y ago
      Ed is my co-founder, he writes all our blog posts because I suck I writing. He also does more than half of our podcast episodes https://thegeomob.com/podcast https://thegeomob.com/podcast (the guy on the left). Last I saw him (yesterday) he was real.
    - freyfogle 4y ago
      Just re-checked the org chart. There's no social media personal assistant.
  - luckylion 4y ago
    I don't understand the original comment to suggest that. Rather: it's a known issue. ChatGPT does nothing new, and certainly doesn't do it by itself -- it just rehashes what others have already written. Like Google might send you visitors for something that's not even present on your website because others link to you mentioning it. What the comment suggested was that they're now bringing this up again to get attention (and links) since it's combined with ChatGPT. That's not "malicious", but it's also not exactly "wow, we just realized this happens".
    - seszett 4y ago
      What the comment suggested is that the company deliberately created tools using their own API in a wrong way in order to write a blog post about it. If that's not an accusation of being malicious I don't know what could be.
vlunkr 4y ago
There's also no clear motive. They want to attract users to a fake feature their free tier?

freyfogle 4y ago

Hi, Ed from OpenCage here, author of the post. We do have python tutorials and SDKs showing how to use our service for ... geocoding, the actual service we provide. I wrote the post mainly to have a page I can point people to when they ask why "it isn't working". Rather than take the user through a tour of past posts I need something simple they will hopefully read. But fair point, I can add a link to last year's post about the erronious youtube tutorials as well. What I think you can't appeciate is the difference of scale. A faulty youtube video drives a few users. In the last weeks ChatGPT is sending us several orders of magnitude more frustrated sign-ups.

singlow 4y ago
I get frustrated at the number of things ChatGPT gets blamed for that aren't its fault. It is completely understandable that if there are repos out on GitHub like the one for Phomber[1] thant ChatGPT would find that code and have no idea that it was phoney. Suggesting that ChatGPT just made this up out of thin air when you know it didn't is not very responsible. 1. https://github.com/s41r4j/phomber https://github.com/s41r4j/phomber
- jraph 4y ago
  You are blaming the victim. OpenAI is to be blamed. They know what they are doing. They provide something that sounds over-confident for anything it says, knowing full well that it can't actually know if what it generated is accurate because it is designed to generate plausible sentences using statistics and probabilities, not verified facts from a database. On top of it, they trained it on an uncontrolled set of texts (though IIUC even a set of verified text would not be enough, nothing guarantees that a LM would produce correct answers). And they provide it to the general population, which doesn't always understand very well how it works and, above all, its limitations. Including developers. Few people actually understand this technology, including myself. Inevitably, it was going to end up causing issues. This post factually presents a problematic situation for the authors of this post. How ChatGPT works or how it can end up producing wrong results is irrelevant to the post's authors problem. It just does, and it causes troubles because of the way OpenAI decided to handle things. And it's not "fair enough, because this false stuff can be found on the internet".

gus_massa 4y ago

That explains why ChatGPT is confused. It may be an old problem, but I guess users are more use to a random YouTube video with wrong information. But the computer is always right so ChatGPT is always right, so users may be more annoyed to discover that the recommendation is wrong and blame them instead of ChatGPT.

fwlr 4y ago

Devs making baby’s first mobile app add “request location information” permissions, the devices start giving them the phone’s GPS information in the form of lat/lon pairs, and those devs naturally look for a service to make that data useful. What they want is “reverse geolocation”, i.e. take a lat/lon pair and return information that makes sense to a human (country, state, nearby street address, etc). This is a service that OpenCage provides, and for whatever reason OpenCage happens to be one of the popular services for this use case. (Maybe it’s because you get the text description of location back right away without having to do a round trip through a heavyweight on-screen map, maybe their free tier allows more requests than most, maybe their api is easier to use, maybe they are lucky or skilled with SEO and their tutorial happens to be the first result for some common phrases, who knows.) So there’s this process that starts with a search for “convert phone location to address”, often involves the OpenCage api, and ends with a happy developer getting the information they wanted. Various algorithms pick up on the existence and repeated traversal of this happy path. In another part of the internet, code tutorial content farms notice a demand for determining an incoming call’s location from the number that’s calling. They search for things like “convert phone number to location” and “convert phone number to address”. Some of these searches end up falling into the nearby well-trodden path of “convert phone location to address” and the content farmer is presented with the OpenCage api. They mess around with the api for a bit and find they can start from a phone number and get a successful api call that returns a lat/lon pair. A successful api call that returns legitimate-looking lat/lon data is all they need to make a video, they make it and post it. Higher-quality, more scrupulous code tutorials attempt to answer this same demand but find it’s not possible, so those tutorials don’t get made, leaving the less scrupulous ones that stop with a successful-looking api call to flourish in this space. The tutorial is doing well, so the content farms endlessly recycle it into blogspam. As a result, OpenCage starts getting weird usage patterns, tracks them down, finds the source is these tutorials, and makes a post about it. Some time later, ChatGPT is released. People are astounded with its ability to write code and start using it for this purpose. Naturally, some of those people have the same demand as the previous generation of devs who stumbled onto the unscrupulous code tutorials. Because of the blogspam, ChatGPT’s training data includes many variations on the tutorial, and just as naturally it ends up reproducing that tutorial when asked - except ChatGPT’s magic kicks in and instead of including (what its embeddings see as) some weird unrelated area-code-to-string nonsense from the tutorial, it just bullshits some plausible-sounding data plumbing code instead. Unfortunately, because the tutorial never worked in the first place, that weird hacky irrelevant bit that ChatGPT ignored happened to be the secret sauce that makes the whole thing superficially appear to work. As a result, OpenCage starts getting weird usage patterns, tracks them down, finds the source is ChatGPT, and makes a post about it. In deference to Hacker News’ policy of keeping comments pleasant, I will elide the analysis of the process that leads to comments accusing OpenCage of nefariously engineering the whole thing for attention.