Hacker News Reader

Comment by tsunamifury | Hacker News Reader

This sounds a bit like word salad, and your argument is worse than an LLM output. You even contradict yourself at the end.

Also I think you need to better understand how much hallucination has been driven down recently and its clear path going forward.

View on HN

I just asked ChatGPT to give me directions to my favorite shop in Manhattan from Penn Station. It gave me wrong public transit directions about which subway to take. It also gave me wrong walking directions, putting me blocks off course and as to which side of the street I'd find the shop on.

The only thing it got right was: "Please note that subway schedules and routes can vary, so it's a good idea to use a navigation app like Google Maps or the official MTA website for real-time information and step-by-step directions."

Well I guess that ends the debate then.

ChatGPT: not even as good as MapQuest.

Checkmate, OpenAI.

yea there seems to be such a weirdly deliberate way to use it incorrectly.

I absolutely agree, and I've been in several similar situations. Yesterday I finally caved and gave it very clear instructions on how I wanted it to build my new IKEA shelf, even though understanding the instructions is most of the work, and guess what? The box is still sitting there, unopened! It hasn't even started. I asked it if it could fight Malenia for me, same thing, more useless machine babble. Repair the slipping rubber on my office chair, incoherent results and not even an attempt to start. Factorise a prime number (I just wanted to break the encryption on some traffic I got a hold of) and it didn't work, again. Just now, I asked ChatGPT what number I was thinking of and it got it so unbelievably wrong... I can't even believe people are using this crap. ChatGPT has very much not come good on the promises OpenAI made about what it could do.

This is kind of like if a fisherman walking down the street looked up at a steel sky scraper and proclaimed:

"Gee, how can they be making entire buildings out of that?! Why my plastic dinghy could be out at sea for a year and not pick up a lick of rust, meanwhile my steel fishing hooks are rusty if I leave them out on a rainy night!"

I'm really unsure if you are making a joke, but in case you're serious: using a language model for directions is like using a thermometer for calculations. Sure, it uses numbers...

The problem is that actual people are trusting the output they get, not realizing that it's just plausible sounding, not correct. It is surprising just how plausible it sounds. Which is the uncanny valley into which any number of people will walk off the proverbial cliff and fall into. The hype surrounding the capabilities of LLMs far exceeds its actual output.

People trust all kinds of bad output. Look at politics and how hyped people get over the promises that are never kept.

> Also I think you need to better understand how much hallucination has been driven down recently and its clear path going forward.

Unless it’s essentially fully eliminated, it doesn’t qualitatively change the argument. The output cannot be trusted to go directly to customers and/or business partners. It can’t even be trusted to go to employees that don’t know it’s from AI and might lie if you ask if something strange.

> You even contradict yourself at the end.

There is no contradiction. Read it again.

“This isn’t fixable,” said Emily Bender, a linguistics professor and director of the University of Washington’s Computational Linguistics Laboratory. “It’s inherent in the mismatch between the technology and the proposed use cases.”

https://apnews.com/article/artificial-intelligence-hallucina...

Bender also recently said in some podcast that there's basically no useful case for LLMs. I think she's a good candidate for some soundbite to add to "a global market of maybe five computers", "nobody needs more than 640k", etc (but she is probably not notable enough even for that).

Reply on news.ycombinator.com