[simias]

I haven't tried ChatGPT for coding yet but I have used it to study human languages (I've been studying Japanese for a little while) and it's so easy to get it to spew complete nonsense with perfect aplomb that it makes me super wary of using it for any "serious" things. In particular anything you're not super familiar with where you can't easily evaluate if the output is correct or not.

The other day I was struggling to parse a Japanese sentence, a particular grammatical construction made no sense to me. I wrote the sentence in ChatGPT, asked it to break it down for me, and it came up with a plausible-sounding explanation. Problem was, I couldn't find any hit on google when I searched for the thing it was talking about. So I asked ChatGPT to give me more details, tell me what I could search for, and it would insist that its explanation was correct and then gaslight me by telling me that the reason I couldn't find anything on Google was because it was a niche subject not usually taught in grammar books.

After some more searching around and double-checking it turns out that I had misread a kanji and the sentence I typed into ChatGPT was complete gibberish as a result. ChatGPT's explanation, while sounding very plausible, was complete fabrication.

The idea that some inexperienced people are shipping software using this tool is insane to me.

[gertlex]

> The idea that some inexperienced people are shipping software using this tool is insane to me.

I really think you're trying to compare apples and oranges, in multiple ways. For one, we can test the software by running it, which is a pretty different problem from asking language questions, with a much slower ability to verify correctness (based on what you describe and what I imagine).

I'm not saying your experience is invalid. In my own adventures, the equivalent of what you did was my writing some incomplete bash in an existing script, wandered off to another part of the code. I then came back to that incomplete snippet, and though it was some unfamiliar syntax someone else had written (vim even highlighted it like it was special!). Naturally I went and asked ChatGPT what that snippet did, and wasted 15 minutes trying to corroborate it before checking the git history or something and realizing my own error.

[simias]

>For one, we can test the software by running it

As long as the tests are not also written by ChatGPT...

Many critical security issues require a deep understanding or the code or some intense fuzzing to discover, it's not enough to ask ChatGPT "write me X" then superficially glance at the output to validate that it looks correct. That's the part that worries me. Completely broken code will be caught immediately, but subtly broken code may linger for a long time and make it to production.

And from my limited experience with ChatGPT, it seems very good at making up broken things that look superficially correct.

[renonce]

Just wondering, are you trying it with GPT-3.5 or GPT-4? Better to use GPT-4 API version since the web version’s quality was controversial. When you feel quality is subpar you should try to use the best one first.

[simias]

I'm not sure but given that I haven't paid anything I believe it's 3.5.

[imposter]

Maybe they didn't feed it enough linguistical data for japanese, but they surely did for code.

[ed_elliott_asc]

The thing that annoys me is that you can ask it to do something specific but it does what it wants.

For example I had a massive sql query that was loads of statements unioned together, I said “for each statement remove this filter and add this filter” and it would go “certainly, here are the first four, I have used an example table name feel free to change it” then I’d say “can you do it for all the statements, not just the first four” and it would say “of course here you go” and just give me the first four but also makes them useless by changing the table name!

I’ve got great hopes that one day I can get it to help me shape the code in a way that the jetbrains ide’s can’t today - for those I have to choose from a set of available operations - I want to talk to it and get it to change the code in a set of operations that I choose!

One day maybe :)

[simias]

Perhaps, but I wouldn't mind if the models just answered "I'm sorry but I don't have an answer to your question at the time". In fact I think that would be a great answer that would increase the amount of trust I have in ChatGPT.

Instead the model decides to make stuff up and pretend that it knows. That's vastly worse.

It reminds me of the early days of DuckDuckGo, when if you searched for something obscure with no matches online it would still fuzzy match some garbage like a binary blob in a Chinese PDF while Google helpfully would just tell you that it couldn't find anything.

[Cannabat]

> Instead the model decides to make stuff up and pretend that it knows. That's vastly worse.

Does the model know it doesn’t know though? Does “know” even make sense as a concept here? I don’t know if it can really introspect like that, but of course it would be so much better if it could can have some sort of confidence score with each answer.