Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Don't believe ChatGPT – we do not offer a "phone lookup" service

blog.opencagedata.com

Post reply on HN
Print view View original

101–110 of 284 posts

Re: Don't believe ChatGPT – we do not offer a "phone lookup" service

#101

Post by pavon »

mochomocha wrote: post #63
crazygringo wrote: post #54

I'm curious -- does anyone know of ML directions that could add any kind of factual confidence level to ChatGPT and similar? We all know now that ChatGPT is just autocomplete on steroids. It produces plausibly convincing patterns of speech. But from the way it's built and trained, it's not like there's even any kind of factual confidence level you could threshold, or anything. The concept of factuality doesn't exist…

> But from the way it's built and trained, it's not like there's even any kind of factual confidence level you could threshold, or anything. The concept of factuality doesn't exist in the model at all. I'm not super familiar with ChatGPT internals, but there are plenty of ways to tack on uncertainty estimates to predictions of typical "large scale ML models" without touching Bayesian stuff (which only work for small…

Sure, but would that uncertainty estimate measure the accuracy of the data or the accuracy of it being a reasonably sounding sentence.

Re: Don't believe ChatGPT – we do not offer a "phone lookup" service

#102

Post by shawntan »

snowstormsun wrote: post #62
crazygringo wrote: post #54

I'm curious -- does anyone know of ML directions that could add any kind of factual confidence level to ChatGPT and similar? We all know now that ChatGPT is just autocomplete on steroids. It produces plausibly convincing patterns of speech. But from the way it's built and trained, it's not like there's even any kind of factual confidence level you could threshold, or anything. The concept of factuality doesn't exist…

I think "Explainable AI" is a related research direction, but perhaps not popular for language models.

I think part of the issue is what level of explanation is satisfactory. We can explain how every linear transformation computes its output, but the sum of it is in many ways more than its parts.

Then there are efforts that look like this one: https://news.ycombinator.com/item?id=34821414 They go probing for specific capabilities of Transformers to figure out which cell fires under some specific stimulus. But think a little bit more about what people might want from explainability and you quickly find that something like this is insufficient.

There may be a tradeoff we're looking at where explainability (for some definition of it) will have to be exchanged for performance (under some set of tasks). You can build more interpretable models these days, but you usually pay for it in terms of how well you do on benchmarks.

Re: Don't believe ChatGPT – we do not offer a "phone lookup" service

#103

Post by shagie »

insane_dreamer wrote: post #77
mort96 wrote: post #15

That's quite the predicament. I hope OpenAI will listen, to this and to anyone else in a similar situation. I'm reminded of the cases of ChatGPT recommends random people's personal phone numbers for various services. But yeah, don't trust ChatGPT for anything. Just earlier today I tried my darnedest to convince it that 2 pounds of feathers doesn't weigh the same as 1 pound of bricks, and it just would not listen, pre…

> don't trust ChatGPT for anything Agreed. But then it begs the question: what purpose does ChatGPT serve (other than for entertainment purposes or cheating on your HS/college exam)? If you have to verify its information by other means, then you're not really saving much effort.

It works really well for translating one "language" to another "language".

Give it some structured data and ask it to summarize it (e.g. hourly weather data and it gives a better summarization than a template based one).

Give it HN titles and the categories and it does a passable zero shot tagging of them ( https://news.ycombinator.com/item?id=34156626 ).

I'm toying around with making a "guided bedtime story generator". A friend of mine uses it to create a "day in the life of a dinosaur" stories for a child (a different story each day!)

The key is to play to its strengths rather than testing its bounds and complaining that they break in weird ways when they will inevitably break in weird ways.

Re: Don't believe ChatGPT – we do not offer a "phone lookup" service

#104

Post by wvenable »

VectorLock wrote: post #88

This is the biggest problem I encounter when trying to use ChatGPT on a daily basis for computer programming tasks. It "hallucinates" plausible looking code that never existed or would never work, especially confusing whats in one module or API for something in another. This is where ChatGPT breaks when pushed a bit further than "make customized StackOverflow snippets." For example I asked ChatGPT to show me how to u…

The problem is compounded by the fact that sometimes it produces really good results. One task, good results. Next task, totally hallucinated result.

Re: Don't believe ChatGPT – we do not offer a "phone lookup" service

#105

Post by int_19h »

yieldcrv wrote: post #4

lol it recommended their api and gave python code for using it but the real api doesnt give results that the user asked ChatGPT for that is amusingly alarming

The obvious follow-up is to create the non-existing API endpoint but hook it into GPT so that it can hallucinate a convincing address based on the phone number. Take GPT API key as input so that the caller is paying for this.

Bonus points for using ChatGPT to implement this end-to-end.

Re: Don't believe ChatGPT – we do not offer a "phone lookup" service

#106

Post by amscanne »

irrational wrote: post #87
crazygringo wrote: post #54

I'm curious -- does anyone know of ML directions that could add any kind of factual confidence level to ChatGPT and similar? We all know now that ChatGPT is just autocomplete on steroids. It produces plausibly convincing patterns of speech. But from the way it's built and trained, it's not like there's even any kind of factual confidence level you could threshold, or anything. The concept of factuality doesn't exist…

Remember the guy a few weeks ago that was being gaslighted by ChatGPT that this is the year 2022? Not only is it giving out potentially false info, but it will double down that it is right and you are wrong. Though, to be honest, that sounds like a lot of real people. The difference is, people are smart enough to not double down on try to say it is a different year and your phone is probably reporting the year wrong.

That was the Bing preview, which is supposed to be an actual information product.

Re: Don't believe ChatGPT – we do not offer a "phone lookup" service

#107

Post by coldtea »

circuit10 wrote: post #91
coldtea wrote: post #24

ChatGPT doesn't "recommended" anything. It just recombines text based on statistical inferences that appear like a recommendation. It could just as well state that humans have 3 legs depending on its training set and/or time of day. In fact it has said similar BS.

> ChatGPT doesn't "recommended" anything. It just recombines text based on statistical inferences that appear like a recommendation. I think that’s a bit pedantic and not very helpful… I’m not typing this comment, my brain is just sending signals to my hands which causes them into input data into a device that displays pixels that look like a comment

>I think that’s a bit pedantic and not very helpful… I’m not typing this comment, my brain is just sending signals to my hands which causes them into input data into a device that displays pixels that look like a comment

Well, if you're just fed a corpus, with no real-time first-person strem of experience that you control, no feedback mechanism, no higher level facilities, and you're not a member of a species with a proven track record of state-of-the-art in nature semantic understanding, then maybe...

Re: Don't believe ChatGPT – we do not offer a "phone lookup" service

#108

Post by coldtea »

mort96 wrote: post #32
coldtea wrote: post #24

ChatGPT doesn't "recommended" anything. It just recombines text based on statistical inferences that appear like a recommendation. It could just as well state that humans have 3 legs depending on its training set and/or time of day. In fact it has said similar BS.

What would you call it instead?

A glorified Markov chain generator.

Now, humans could very well also be statistical inference machines. But they have way more tricks up their semantic-level understanding sleeves than ChatGPT circa 2023.

Re: Don't believe ChatGPT – we do not offer a "phone lookup" service

#109

Post by elicash »

ntonozzi wrote: post #53

Including the word 'phone' six times in a popular blog post is not going to help their predicament.

Wouldn't they want this post to be at the top when people search 'phone' and 'open cage data'? Seems like SEO towards correcting this is only helpful. And maybe when GPT updates data, this post gets pulled in, too. The more popular, the better, I'd guess.

Re: Don't believe ChatGPT – we do not offer a "phone lookup" service

#110

Post by mtmail »

IshKebab wrote: post #41

Well for a start you could make it more obvious what your service does do. I don't know what "geocoding" is. Converting things to/from "text" is meaningless. You have to get all the way down ... way down, past authentication to the details of the `q` query parameter before it actually tells you. At the top you should have a diagram like this: Lat, lon address With a few examples underneath.

"Past authentication", so you're looking at the https://opencagedata.com/api page. Most people go to the homepage first. Great feedback, we should make it clearer on that page and add examples earlier. Thanks!
Post reply on HN