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

The librarian immediately attempts to sell you a vuvuzela

kaveland.no

Post reply on HN
Print view View original

41–50 of 346 posts

Re: The librarian immediately attempts to sell you a vuvuzela

#41

Post by myself248 »

ars wrote: post #24
sgarland wrote: post #16

Earlier quoted context omitted.

Altavista was the OG. I remember it being cantankerous and requiring you to specific in how you searched, but if you knew how to use it, it was unmatched. Until Google.

When Google came out it was way better than Altavista, people switched instantly. Specifically Altavista looked at how often a search term was in the result, which wasn't always a helpful thing. Google also noticed if search terms were near each other in a page which was really helpful, otherwise you would get forums with one search term in one message, and the other far away in an unrelated message. Google fixed tha…

When Google came out, I started using it for some things, because yes it was better at some things, but I didn't stop using Altavista. Stayed with it until the very end, for cases where I could be certain that I knew the exact words that'd be in the page, and Google just sucked at that.

These days I can't even -exclude terms that I know would only appear in the wrong results, Google will show me those results anyway. Nothing about adversarial SEO requires them to ignore my input, that's a different choice.

Re: The librarian immediately attempts to sell you a vuvuzela

#42

Post by WesolyKubeczek »

moron4hire wrote: post #7
boothby wrote: post #2

> These days, I find that I am using multiple search engines and often resort to using an LLM to help me find content. For a few months, I've been wondering: how long until advertisers get their grubby meathooks into the training data? It's trivial to add prompts encouraging product placement, but I would be completely shocked if the big players don't sell out within a year or two, and start biasing the models themse…

It's kind of already happening. For example, if you ask an LLM for advice on building an application, it's going to pigeon-hole you into using React.

That’s because of statistical likelihood and abundance of web content about React which seems to be kinda default choice. Had to be a looong con if it was an ad.

Re: The librarian immediately attempts to sell you a vuvuzela

#43

Post by reubenmorais »

boothby wrote: post #2

> These days, I find that I am using multiple search engines and often resort to using an LLM to help me find content. For a few months, I've been wondering: how long until advertisers get their grubby meathooks into the training data? It's trivial to add prompts encouraging product placement, but I would be completely shocked if the big players don't sell out within a year or two, and start biasing the models themse…

Google has been working on auctioning token-level influence during LLM generation for years now: https://research.google/blog/mechanism-design-for-large-lang...

Re: The librarian immediately attempts to sell you a vuvuzela

#44

Post by galaxyLogic »

hug wrote: post #15

The author would do well to check out Kagi -- the search results for all of the suggested queries are, to my eye, better than the results they found. Given that Kagi's higher tier plans come with search-enabled LLM chat interfaces, and those searches use Kagi's results (which, again, appear to be superior) it seems to me that you get the best of both worlds: Better search, and better search results to feed into your…

I've been using Preplexity lately while learning to use Yamaha SeqTrack synthesizer. I have the UserGuide, but I find it easier and faster to ask Preplexity for things like "How do I mute a track?".

It never occurred to me I could use Google this way. And it is a novel idea to me, that it seems to be better to use AI than read a manual.

Re: The librarian immediately attempts to sell you a vuvuzela

#45

Post by jedbrooke »

boothby wrote: post #2

> These days, I find that I am using multiple search engines and often resort to using an LLM to help me find content. For a few months, I've been wondering: how long until advertisers get their grubby meathooks into the training data? It's trivial to add prompts encouraging product placement, but I would be completely shocked if the big players don't sell out within a year or two, and start biasing the models themse…

not quite ads in LLMS, but I had an interesting experience with google maps the other day. the directions voice said "in 100 feet, turn left at the ". Normally it would say "at the traffic light" or similar. And this wasn't some easy to miss hidden street, it was just a normal intersection. I can only hope they aren't changing the routes yet to make you drive by the highest bidder

Re: The librarian immediately attempts to sell you a vuvuzela

#46

Post by Peteragain »

A proposal for a solution. The data is the unique selling point. Put that in public hands with an API, published algorithms, and it's own development team. The free market can then sell user interfaces, filters, and whatever. The metaphor is roads (state managed) and vehicles (for profit). Today I can (physically) go to the British Library and get any published book, or go on line and pay for the privilege.

Re: The librarian immediately attempts to sell you a vuvuzela

#47

Post by eviks »

> These days, I find that I am using multiple search engines and often resort to using an LLM to help me find content.

> This is just my anecdata

Not a single example, so it's not even that, just vibes

> the past few years. There’s recent science to read about the quality of search.

Let's look at the "science"

> We monitored Google, Bing and DuckDuckGo for a year on 7,392 product review queries.

Oh, so there is nothing about the past few years either. And this isn't "search", but a very narrow category of search that is one of the prime targets for SEO scam, so it's always been bad, and the same incentives made review content highly suspect before any SEO was ever involved.

Which multiple engines can help you here? Which LLMs?

Re: The librarian immediately attempts to sell you a vuvuzela

#48

Post by petesergeant »

I mean sure, but also, I feel like the ability to query an LLM for something is an invaluable resource I never had before and has made knowledge acquisition immeasurably easier for me. I definitely search the web much, much less when I'm trying to learn something.

Re: The librarian immediately attempts to sell you a vuvuzela

#49

Post by collingreen »

jedbrooke wrote: post #45
boothby wrote: post #2

> These days, I find that I am using multiple search engines and often resort to using an LLM to help me find content. For a few months, I've been wondering: how long until advertisers get their grubby meathooks into the training data? It's trivial to add prompts encouraging product placement, but I would be completely shocked if the big players don't sell out within a year or two, and start biasing the models themse…

not quite ads in LLMS, but I had an interesting experience with google maps the other day. the directions voice said "in 100 feet, turn left at the ". Normally it would say "at the traffic light" or similar. And this wasn't some easy to miss hidden street, it was just a normal intersection. I can only hope they aren't changing the routes yet to make you drive by the highest bidder

This is devilish. I'm adding your idea to my torment nexus list.

Re: The librarian immediately attempts to sell you a vuvuzela

#50

Post by vintermann »

bdangubic wrote: post #22
iamkeithmccoy wrote: post #20

> AIs are much better at responding to my intent, and they rarely attempt to sell me anything Yet. It's only a matter of time before AI becomes ad-riddled and enshittified.

they won’t as long as you are paying $200+/month

Bold to assume they won't try to double-dip and profit from you in more than one way. Cable TV got ads, subscription newspapers got ads (and advertorials). Just because it's not free doesn't mean you aren't the product.
Post reply on HN