Generative AI could make search harder to trust • Hacker News Board

Re: Generative AI could make search harder to trust

#251

Post by darkerside » Sat, Oct 07, 2023, 4:28 PM UTC

TerrifiedMouse wrote: post #243

darkerside wrote: post #238

Earlier quoted context omitted.

Again... > Because everyone has at least tried to lie - not necessarily for malicious reasons, e.g. white lies - a few times in their lives and know it’s not easy coming up with plausible sounding lies. How do you know that?

> How do you know that? Common knowledge. Regardless, the amount of false information on the internet used to be limited by the amount of people producing it. Some of it is deliberately produced misinformation. Some of it are just mistakes due to carelessness, others due to willful negligence - example of the latter would be content farms who make their money off flooding search engine result pages and pushing ads in…

I guess I just disagree on either how capable people are of thinking on their feet, or what your definition of bullshit is. Still not sure which.

Re: Generative AI could make search harder to trust

#252

Post by myaccountonhn » Sat, Oct 07, 2023, 7:23 PM UTC

yoyohello13 wrote: post #178

lhl wrote: post #134

Earlier quoted context omitted.

I had a similar experience recently looking some stuff for Starfield. Content farms are obviously switching to LLM-generated (mostly hallucinated) articles and Google seems to be ranking them pretty highly atm. Kagi's ability to manually downrank/remove those kinds of results from your searches (and their return to flat rate pricing) finally tipped the scales for me for subscribing/switching search.

I actually found my first AI YouTube channel today looking for starfield videos. I noticed the narrator sounded like text-to-speech. I went to the channel and it’s all just random videos for tones of different games with no real theme. Was a surreal experience.

I had mine the other day when I bought an audiobook and noticed the price was way cheaper and sure enough, it was a very robotic voice with uncanny pauses reading it. Refunded instantly

Re: Generative AI could make search harder to trust

#253

Post by TerrifiedMouse » Sat, Oct 07, 2023, 11:20 PM UTC

darkerside wrote: post #251

TerrifiedMouse wrote: post #243

Earlier quoted context omitted.

> How do you know that? Common knowledge. Regardless, the amount of false information on the internet used to be limited by the amount of people producing it. Some of it is deliberately produced misinformation. Some of it are just mistakes due to carelessness, others due to willful negligence - example of the latter would be content farms who make their money off flooding search engine result pages and pushing ads in…

I guess I just disagree on either how capable people are of thinking on their feet, or what your definition of bullshit is. Still not sure which.

Well, you can be sure the SEO creating content farms will pick the path of least resistance.

They won't make up a lie when telling the truth is far easier and when they do have to lie, it's a slight bit more effort.

But that's a moot point with LLMs ...

If/When they use LLMs, they will pick the cheapest LLM possible which will produce the most garbage - they might not even bother keeping it up to date; why bother when hallucinated BS sound just as convincing.

It has already started to happen.

Re: Generative AI could make search harder to trust

#254

Post by p0w3n3d » Sun, Oct 08, 2023, 9:19 PM UTC

euroderf wrote: post #32

p0w3n3d wrote: post #26

And entropy rises... people thought AI will kill us with machine guns. AI will kill us by making us super stupid...

I have already externalized my to-do lists and other reminder lists to teh interwebz. I can't wait to outsource my faculties for reasoning too.

To be honest I really am scared how quickly my mind removes some ability once it gets automated. I already almost stopped looking into mirror while rear parking because one of my cars has rear parking sensors. I'm scared of driving a car with front parking sensors...

Re: Generative AI could make search harder to trust

#255

Post by PaulHoule » Fri, Oct 13, 2023, 3:59 PM UTC

StableAlkyne wrote: post #127

I actually experienced this the other day. Bought the new Baldur's Gate and was wondering what items to keep or sell (don't judge me, I'm a pack rat in games!) I had found some silver ingots. The top search result for "bg3 silver ingot" is a content farm article that very confidently claims you can use them at a workbench in Act 3 to upgrade your weapons. Except this is a complete fabrication: silver ingots exist onl…

10 years ago I was a huge fan of GameFAQs, which, for any major game, had at least one and often several highly detailed documents describing a game, not just a walkthrough but a set of tables where you would find items.

Back then I was playing Hyperdimension Neptunia and almost tried “applying AI” in the old sense to the problem of “What do I have to do to craft item X?”. Those games all had good FAQs and extracting a knowledge graph and feeding it into some engine that could resolve dependencies wouldn’t have been too hard.

Today I am playing Atelier Sophie which has the same mechanic but is very cozy and doesn’t pose complex dependency problems and the FAQs for this game are atrocious, consisting of a walkthrough that way too prescriptive. If you ask some question like “Where do I get a Night Crystal?” on Google this is likely to turn up a question/answer pair on a forum which isn’t quite as good as having the structured FAQ.

YouTube walkthroughs really seemed to kill text walkthroughs, sometimes these are better (like when there is a jump in a dungeon that doesn’t look like you could make it but you can) but sometime they are much worse (there are 75 hour long videos in a walkthrough, you have to find that it is in video #33 and that you have to seek to 20:55.)

Maybe the proliferation of trash sites will motivate the creation of high quality FAQs but you’d better believe that the creator of these FAQs will be horribly afraid of being ripped off.