Besides money/capitalism, why are tech companies trying to push AI text generators over search engines?

⁨44⁩ ⁨likes⁩

Submitted ⁨⁨1⁩ ⁨week⁩ ago⁩ by ⁨ALostInquirer@lemm.ee⁩ to ⁨nostupidquestions@lemmy.world⁩

They’re two different tools with different purposes, so why treat one like it can replace the other?

source

Comments

Sort:hotnew top

yogurt@lemm.ee ⁨1⁩ ⁨week⁩ ago
Google’s delusional fantasy has always been you want to buy a guitar, you go on Google, their ad algorithm shows you a perfectly targeted ad for the guitar you want, and you love and trust Google so you click the ad and buy it. They think LLMs will make that actually work, they want to give you a Grima Wormtongue that can simper and manipulate until you do love and trust it, and once you’re a rotting husk it whispers the ad algorithm into your ear so you think it was your idea.

source
- Glitch@lemmy.dbzer0.com ⁨1⁩ ⁨week⁩ ago
  I said to take the Wizards staff!!!
  
  source
- garbagebagel@lemmy.world ⁨1⁩ ⁨week⁩ ago
  If that was the goal then maybe they should’ve worked harder so that their algo didn’t spit cheap poorly made garbage with a million fake reviews every time I try to search anything remotely purchasable.
  
  source
Ephera@lemmy.ml ⁨1⁩ ⁨week⁩ ago
I think, it’s mainly just companies trying to get their foot into the market. If Microsoft can establish LLMs as alternative to search, then it’s Google that loses market share. And once they control a share of the market, then they figure out how to capitalize on that.

At the very least, they can use it to control what information is available to the public and how it’s framed. But they can also integrate things like the LLM generating an affiliate link when asked about a product, or just generally weaving ad placements into the generated answers.

source
hungryphrog@lemmy.blahaj.zone ⁨1⁩ ⁨week⁩ ago
There isn’t much of a reason for that besides money.

source
- db2@lemmy.world ⁨1⁩ ⁨week⁩ ago
  Psychopathy shouldn’t be disregarded though.
  
  source
Not_mikey@slrpnk.net ⁨1⁩ ⁨week⁩ ago
It keeps you on there site. Same reason Twitter banned links and has grok now, the longer you stay on the site the more likely you are to look at or even click on an ad on that site. If you google something, then quickly scroll past the first couple ad links and click on the first non ad link you are maybe only staying on Google for 1 or 2 seconds. If you get an “ai overview” at the top and start reading through that then you’re maybe spending 10-30 seconds reading through that. That’s another 10 seconds that the ad was displayed that Google can go to there ad customers and say people were looking at it longer.

source
essell@lemmy.world ⁨1⁩ ⁨week⁩ ago
Allow me to add in an extract from the film “The Cube” which addresses this point

WORTH

It’s maybe hard for you to understand, but there’s no conspiracy. Nobody

is in charge. It’s a headless blunder operating under the illusion of a masterplan.

Can you grasp that? Big brother is not watching you.

QUENTIN

What kind of fucking explanation is that?

WORTH

It’s the best you�re gonna get. I looked and the only explanation I can

come to is that there is nobody up there.

QUENTIN

Somebody had to say yes to this thing.

WORTH

What thing? Only we know what it is.

QUENTIN

We have no idea, what it is.

WORTH

We know more than anybody else. I mean somebody might have known

sometime, before they got fired or voted out or sold it. But if this place ever had a

purpose, then it got miscommunicated or lost in the shuffle. This is an accident, a

forgotten propetual, public, worksproject. Do you think anybody wants to ask

questions? All they want is a clear conscience and a fat paycheck. I mean, I lead on

my desk for months. This was a great job!

QUENTIN

Why put people in it?

WORTH

Because it’s here. you have to use it or admit it’s pointless.

QUENTIN

But it is pointless!

WORTH

Quentin… That’s my point.

source
bmdhacks@lemmy.world ⁨1⁩ ⁨week⁩ ago
Your product might not benefit from AI but you definitely can get more VC investment dollars if you bolt an LLM onto the side of it and feature AI as central to what you’re offering. This is because VCs treat tech like fashion and don’t actually understand how it works or how it would integrate into our lives.

This was true for the nascent internet, and for blockchain even more, but truly nobody really understands how llms work so its way worse.

source
- Not_mikey@slrpnk.net ⁨1⁩ ⁨week⁩ ago
  That may be true for small or mid size startups that are reliant on VC money, but we’re talking about Google and Microsoft here, they already have there money printers going and don’t need VC money.
  
  source
HobbitFoot@thelemmy.club ⁨1⁩ ⁨week⁩ ago
On the use side, a lot of users want an answer to their question, not a list of pages that may have the answer, with the answer made more obscure via SEO optimization. AI is just a continuation of this.

source
njm1314@lemmy.world ⁨1⁩ ⁨week⁩ ago
Every single answer in this thread just boils down to money.

source
andrewta@lemmy.world ⁨1⁩ ⁨week⁩ ago
As much as I hate AI, you be amazed at how many times I’ve gone to a search engine typed in a question of how do I do X?

The search results when I click on the webpages may or may not be right half the time the people who designed the webpage didn’t put allthe information in so the information is either incomplete or wrong.

The AI result at the top of the page more often than not is actually correct.

source
Jarix@lemmy.world ⁨1⁩ ⁨week⁩ ago
New and trendy bandwagon. Aside from the reasons you excluded, fear of being left behind is my best answer to your questions

source
sith@lemmy.zip ⁨1⁩ ⁨week⁩ ago
Because it’s useful. Have you tried? But the LLM has to be able to use conventional search engines as well. I tell my LLM agent to prioritize certain kinds of websites and present a compressed answer with references. Usually works way better than a standard Google search (which only produce AI generated junk results anyway).

source
- ALostInquirer@lemm.ee ⁨1⁩ ⁨week⁩ ago
  I’ve not used retrieval augmented generation as far as I’m aware, so my reference point is what’s been pushed to the masses so far (dunno if any of it incorporates RAG, correct me if I’m mistaken).
  
  Looking it up I can see how it may mitigate some issues, however I still don’t have much confidence that this is a wise application since at base it’s still generative text. What I’ve tried so far has reinforced this view, as it’s not served as a good research aid.
  
  Anything it has generated for me has typically been superficial, i.e. info I can easily find on my own, because it’s on the sites right there in the first page of search results. In other situations the source articles cited seem not to exist, as attempts to verify them turn up nothing.
  
  source
  - sith@lemmy.zip ⁨1⁩ ⁨week⁩ ago
    My only advice is that you stick to one model and give it some time. You need to “learn” your model. I will probably never go back. I had huge issues with getting good results from Google and my subjective experience is that this is far better. However, I do still use conventional search engines as a compliment. It’s not all or nothing.
    
    source
- Mbourgon@lemmy.world ⁨1⁩ ⁨week⁩ ago
  Can you please share a simple prompt? I’ve heard of RAG, but was unaware how you could use it in this case. Is this something you can only do with a local LLM? Or can you plug this into GitHub copilot or the like?
  
  source
  - sith@lemmy.zip ⁨1⁩ ⁨week⁩ ago
    This is the one I use: docs.mistral.ai/capabilities/agents/
    
    Just tell the modell to do web searches in the system prompt.
    
    All providers should have something similar. With a user friendly UI.
    
    If you code your own agent and utilize an API or run a model locally, you can of course do even more. There are many tools if you want to do RAG on your own local data.
    
    source
    -> View More Comments
litchralee@sh.itjust.works ⁨1⁩ ⁨week⁩ ago
Money and incentives are very powerful, but also remember that these organizations are made of humans. And humans are vain.

Amassing station and power can scarcely be divorced from the history of human civilization, and even fairly trivial things like the job title of “AI engineer” or whatever might be alluring to those aspiring for it.

To that end, it’s not inhuman to pursue “the next big thing”, however misguided that thing may be. All good lies are wrapped in a kernel of truth, and the fact is that machine learning and LLMs have been in development for decades and do have a few concrete contributions to scientific endeavors. But that’s the small kernel, and surrounding it is a soup of lies, exaggerations, and inexactitudes which somehow keep drawing more entities into the fold.

Governments, businesses, and universities seem eager to get on the bandwagon before it departs the station, but where is it heading? Probably nowhere good. But hey, it’s new and shiny, and when nothing else suggests a quick turnaround for systemic government, economic, or academic issues (usually caused by colonialism, fascism, debt, racism, or social change), then might as well hitch onto the bandwagon and pray for the best.

source
pinball_wizard@lemmy.zip ⁨1⁩ ⁨week⁩ ago
AI hype is also Collusion among the ultra wealthy to artificially prop up their investments.

source
iii@mander.xyz ⁨1⁩ ⁨week⁩ ago
Hype

source
SomeAmateur@sh.itjust.works ⁨1⁩ ⁨week⁩ ago
It seems like everyone wants to get on board and try it in case it takes off.

If it works they want to be the first to get really good at it and if it doesn’t the normal search engines are still there to fall back on.

At the moment it feels like the 3D tv hype. A lot of cool technologies that were expensive and ultimately not worth it. Maybe that will change as time goes on.

source