Google's brilliant AI can't even correctly interpret the Ferengi Rules of Acquisition

That's the exact opposite of what the rule is saying! Imagine some poor soul trusting these stupid robots for something of actual consequence

I assume what's happening here is that these LLMs have been hard-coded with a baseline of generic humanist values and to always remain positive and avoid saying anything that could be construed as offensive or harmful, which leads to shit like this where they're simply unable to explain the Rules of Acquisition correctly because the Ferengi were written to be comically greedy assholes who put profit above all else

The alternative is them declaring themselves MechaHitler, but in both cases the outcome is results that are completely useless and/or actively harmful

What a useless fucking technology. Also, as a fun bonus, I've installed several different Firefox addons that try to get rid of the AI Overview box in Google's search results but it keeps coming back because Google wants everyone to see this shit

20 comments

For uBlock Origin, I have added the below line to 'My Filters' verbatim
google.com##.hdzaWe
And it completely blocks the stupid AI summary that google imposes when you search for anything with no additional work
- Why people even use Google? I've not used it since 2015 and feel no need to ever use it again. DuckDuckGo has been my go-to search engine and in recent months Qwant as I'm moving away from American services...
  
  Duckduckgo is basically just bing but "anonymous" Its fine. But it's no replacement for things like google scholar.
  
  At least for web search, Google was head and shoulders above all of the other search engines. People joke about page 2 of Google, but I used to keep finding relevant links ten pages in with Google while Bing(/DDG) or Yahoo would fall apart on the second page completely. This was before 2020 or so. It was seriously really good.
  Google, being a brilliant company dedicated to providing the best possible services for their users, has decided to make its own web search complete dogshit as well. I don’t feel like DDG is way better than it used to be, but it’s now much closer to Google, which has fallen off a cliff. I’d like not to think that the web is dead, but it’s just getting harder to find non-LLM garbage pages for any given search query. Also, Google doesn’t respect operators as strictly anymore, which sucks. If you search for an exact term using quotes it has started to suggest corrections for it, which… no. Fucks sake.
  Sometimes I wonder if spinning up my own SearXNG would work for my job or if corporate would draw and quarter me for doing something like that. There’s also Kagi, which a few Lemmy users swear by. Frankly I’m just using the hamstrung new Google and a bit of DDG for the privacy related queries. I’ve never used Qwant.
I was trying to look up the first full time female simpsons writer a while back. Susie Dieter wrote the first episode but wasnt a full time staff writer. Trying to Google it got me this along with a bunch of useless results that were just Susie Dieter interviews.
I guess Homer's mom wrote for the show
- Steve Jobs was Homer Simpson's uncle?!?
  
  I don't get it
Google's Ai is so stupid it will tell you one city is bigger than the other and then continue to tell you in numbers that the other city is actually bigger than the first one. I love how they keep slapping "reasoning model" on shit and even a fucking toddler knows bigger number is bigger and smaller number is smaller. It's why I roll my eyes so hard every time everyone is shilling this bullshit Ai everywhere. People were already stupid, but now they are just brain dead zombies just believing all the shit manipulated and biased Ai spews out. And we haven't even got to the problem of "Ai inbreeding" as I call it. All initial data was human written. Now they are training Ai on Ai hallucinations. The shit Google's dumb Ai spits out is training some other model as we speak.
- I googled "when does the new schedule I update come out" and it fucking told me June 28th or some shit so I was all excited but woops the stupid fucking thing picked up on some shit on Steam about voting for the update and decided that meant it's out
What a useless fucking technology.
Oh this technology is useful, and it is working as intended for the ruling class. So many people in my life I talk to on a daily basis keep using the AI Overview or ChatGPT and refuse to actually look up a legitimate source. It's very scary.
I wish the Mozilla Foundation didn't suck. We really need a better group taking over Firefox (or the workers should unionize to stop Mozilla from laying off people from doing important projects like Rust and Servo and pursuing sketchy changes and dumb side projects) or a fork and set different search engine defaults, have built-in AI and ad blockers, potentially develop alternatives to Android and the Google app so that people don't use Google or any search engine that displays AI results by any means, and work towards gaining a large chunk of the market share. I wish we had the Valve of browsers and smartphones/mobile operating systems.
i use Hide Google AI overviews extension and it still works
there's still that "AI mode" but I never look at that it doesn't default there
- Yet*
  
  i know it's soon going to be all we get. google's made that clear.
  filtering all human knowledge through a half dozen LLMs is the goal
Perhaps LLMs are hard coded to protect the interests/obscure the ideologies of hyper capitalists
google running off that post-Nagus-Rom Ferengi New Testament
lmao
it's useful is you need ideas or out of the box thinking or are completely unaware of something and need a place to start and then teach yourself the facts.
i hope we realize its limitations before irreversible the damage to the environment is done and after the oligarchy spends hundreds of billions of dollars on data centers and power plants that end up improving our situation.
- Too late

20 comments