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

ChatGPT: The Slickest Con Artist of All Time

tedgioia.substack.com

Post reply on HN
Print view View original

311–320 of 444 posts

Re: ChatGPT: The Slickest Con Artist of All Time

#311

Post by freejazz »

worik wrote: post #307
freejazz wrote: post #303

Earlier quoted context omitted.

Perhaps you have issues with reading comprehension? This is a thread about how chatGPT is being sold as a service to analyze legal documents, and it quite obviously fails at that. If your solution is to see a lawyer you are making my point that chatGPT is not helpful for this thing that people are saying chatGPT is helpful for.

> Perhaps you have issues with reading comprehension? This is a thread about how chatGPT is being sold as a service to analyze legal documents No it is not

Certainly my posts were and it's a mystery as to what point you think you are achieving by trying to debate something with me that I was never discussing

Re: ChatGPT: The Slickest Con Artist of All Time

#312

Post by amrangaye »

fnordpiglet wrote: post #17

I had a detailed conversion with chatGPT about how to gracefully handle terminating conditions of a rust program. It summarized cogently to register at_exit()’s for each thread, panic handlers, and register signal handlers. It advised and explain in detail on my query about the thread handling for each of these variants, gave really helpful advice on collecting join handles in a closure on the main thread and waiting…

I’ve had the same experience with Python code. I found this article poorly written and alarmist - he really doesn’t prove his main point beyond hand waving, and phrases about MS about to “hand over” their search engine to the AI sounds like someone who really doesn’t know much about GPT. Guess I’m disappointed because when I clicked the title I was expecting a cogent criticism I could learn from.

Re: ChatGPT: The Slickest Con Artist of All Time

#313

Post by amrangaye »

cribbles wrote: post #58

I would greatly appreciate a moratorium on this genre of article until there is compelling accompanying evidence that a meaningful portion of ChatGPT's users are unaware of these shortcomings. I have yet to encounter or even hear of a non-technical person playing around with ChatGPT without stumbling into the type of confidently-stated absurdities and half-truths displayed in this article, and embracing that as a lim…

+1. Not even sure this is “eligible” for an hn post. It actually makes less sense than the CNN ones I saw earlier, and boy those were terrible takes.

Re: ChatGPT: The Slickest Con Artist of All Time

#314

Post by ilteris »

danielbln wrote: post #228
jwie wrote: post #67

Earlier quoted context omitted.

It’s only as good as the information it harvested. In a space where less true information is available, the results would be quite poor.

There are already ways of dynamically populating the context of the model with information retrieves as text embeddings from a vector store, based on the initial prompt. This way you can feed the model current and correct information which it will weave into its output

Do you have an example or a tutorial of this please?

Re: ChatGPT: The Slickest Con Artist of All Time

#315

Post by IshKebab »

ResearchCode wrote: post #26
ajuc wrote: post #7

"Cars won't replace horses, because they require roads and horses don't."

Still waiting on those autonomous cars. Unlike many other software development tools, I've found ChatGPT to be useless so far.

They clearly won't replace autonomous horses.

Re: ChatGPT: The Slickest Con Artist of All Time

#316

Post by gunshai »

bsaul wrote: post #288
gunshai wrote: post #49

My stance is pretty simple. The folks that adapt their own language centers and domain reasoning around using chatGPT (or these types of models) will stand to gain the most out of using them. This article is an eye roll to me, a calculator gives you confidence as well, doesn't mean you used it correctly. It is very hard for me to not outright dismiss articles like this that don't consider the usefulness of the tool.…

You're confounding the problem of feeding a system with incorrect inputs ( calculator ), with a system that, provided the input are correct, will throw garbage totally at random. This is a very different kind of system with very different kinds of problems caused to its users.

I think my caveat of domain knowledge side steps this issue you bring up. It's the entire reason engineering schools teach the "long way" when you're going to rely on the "short way" 99.99% of the time.

Also the same could be true for digital calculators when first created. The neural nets (aka humans) that were initially trained to design the system can take short cuts for example on improperly returning or applying functions such as square roots ect. It still confidently gives you the incorrect answer. Given sufficient training data (IE the human) the system improves to return better more accurate results.

Re: ChatGPT: The Slickest Con Artist of All Time

#317

Post by IshKebab »

dougmwne wrote: post #12

Pretty cool that GTP is hitting such a mainstream moment. Everyone I talk with about it has glazed over for years, but I guess this is finally a demo that breaks through. 100m users if reports are accurate. Of course regular folks are going to wildly overestimate GTP’s current capabilities. Regular folks wildly overestimate the intelligence of their pets.

Yeah this is not really news to anyone that has played with ChatGPT or understands how it was trained. It was optimised for plausibility not truth. Of course it's going to bullshit.

I'm sure there are many smart people working on making it bullshit less, and I don't see any reason to think it is an insurmountable problem.

Also there are many many useful applications where the bullshit doesn't really matter that much.

Re: ChatGPT: The Slickest Con Artist of All Time

#318

Post by nr2x »

Reptur wrote: post #296
qayxc wrote: post #275

Earlier quoted context omitted.

> Took 20-30 follow-up messages, telling it to add things, remove things, fix things, make things compatible with the versions of software it was running etc. So in other words you basically spent just as much time and effort as if you did it yourself?

I understand your point, and you are right. Had I been familiar with the image I was working with, and the version differences in configuration etc. It probably would have taken me the same amount of time. I look at this from a more zoomed out perspective, this is just the beginning. The point is its capable and will improve.

Exactly.

Re: ChatGPT: The Slickest Con Artist of All Time

#319

Post by HarHarVeryFunny »

speed_spread wrote: post #245
pdonis wrote: post #78

Earlier quoted context omitted.

> It is basically like having an incredibly smart engineer/scientists/philosopher/etc that can explain things quite well, but for pretty much every field. No. You've fallen for exactly the con the article is describing. ChatGPT knows nothing . It is very good at generating text that seems plausible if you know nothing about the subject matter, but is not reliable as an actual source of information . If you happened t…

There's lots of real knowledge encoded in ChatGPT and it can recombine and regurgitate it in enough ways that I'd be okay saying that it knows stuff. The real problem is that it doesn't know its limits and will make shit up instead of saying "I don't know". How to reliably detect and enforce these limits probably requires a completely different approach to AI.

I think ultimately any truly intelligent entity needs to have agency - to be able to interact with the world and test truth (prediction correctness) for itself. Without this you're really just an expert system - a bunch of canned data/rules.

The strength of ChatGPT over a GOFAI expert system like Cyc is how broad it's knowledge base is, but the weaknesses are that is doesn't know whether this "knowledge" (being self-deduced from the training set) is true or not, nor does it have any reliable way to combine facts/sources, since those rules are also self-derived and unverified. I'm sure some of this could be mitigated by more human curation (which seems to be the path OpenAI is going down), but then it seems it is just an expert system (maybe this is all people are looking for anyway).

Re: ChatGPT: The Slickest Con Artist of All Time

#320

Post by nr2x »

fnordpiglet wrote: post #17

I had a detailed conversion with chatGPT about how to gracefully handle terminating conditions of a rust program. It summarized cogently to register at_exit()’s for each thread, panic handlers, and register signal handlers. It advised and explain in detail on my query about the thread handling for each of these variants, gave really helpful advice on collecting join handles in a closure on the main thread and waiting…

Yup, told it I was trying to build a search interface on top of a huge Postgres database and wanted ways to speed it up. Told me to check out Elasticsearch, which is promising.
Post reply on HN