Many people I know are also using LLMs for searching stuff.
Its just baffling how they can trust a single answer frankensteind together from often unknown sources presented as a fact.
Jeez we are fucked, aren’t we?
My 10 yr old figured it out when googles AI couldn’t tell the difference between spiderman: miles Morales and spiderman 2. “Why is it trying to give me trophies for the old game?”
That’s a great lesson for them though! If it’s bad at something you’re familiar with, what else does it have wrong?
Me when podcasters I like start talking about a subject I’m very informed in
I miss back when podcasts were just folks pissing about, I remember listening to one back in middle school where it was just a couple of guys talking about weird shit they found on the internet while slowly getting increasingly shit faced.
At least it beats googles results nowadays. At least in productivity. Instead of trying to figure out which results were vaguely congruent with my search terms and not just SEO-shit, I get shitty results instantly. And with the confidence of absolute certainty that 2 prime dunning-kruger-examples would show.
Clear winner 😁
ChatGPT to Google/Wiki isn’t a bad workflow for factual questions.
Nono, I didn’t want to condemn LLMs. They do have their areas where they really shine. As long as you don’t mindlessly just accept everything as fact. I use them all to a good amount on a regular basis.
Dead Internet theory is becoming more and more plausible by the day.
Doesn’t ChatGPT also use google?
I tested this whole concept with Mistral AI. It searches the web, aggregates its findings and provides an answer highlighting potential perspectives / different answers with each one providing a link to the source URL. As much as I hate AI, it does work great that way (since the LLM doesn’t have to pull stuff out of its butt).
Are you certain that the answer was actually from their sources? I had multiple occasions where Mistral/ChatGPT gave me sources and I felt like something was off. I then followed the sources and could not find what they found according to themselves. I then asked them to quote the actual text they used to provide said answer and after drilling them a few more times they concluded that yes, the thing they said was actually not anywhere to be found in the sources they provided.
Even if they weren’t from the source, if it adds a link you can avoid hallucinations by reading the original website it supposedly found the information on.
I know this works with Perplexity.ai and the paid version of ChatGPT
This has also happened to me several times.
Mmh, perhaps I lucked out or missed something. Everything looked good when I tested it.
Breathtakingly fucked.
2 weeks ago I had an issue with my hardware Steam Link setup. For some reason picture would freeze but sound would still go. It is connected over ethernet so I ruled out connection issues. Googling lead me to nowhere. Adding “reddit” to search had no effect. Then I just decided to try chatgpt. It gave me few things I can try and lo and behold, second suggestion was what I was looking for.
There are things I would rather not ask LLMs. But when I have no clue what to even write in google search bar, I’d rather go for LLM because it can lead me to what I am searching for and from there I can continue my own research.
yeah i feel that, but i also feel the guilt of using 3 glasses of water to cool down the computer that answered me.
at any rate, please do not use grok.
elon is cooling his penis computer with aquifer drinking water from the memphis sands aquifer.
he promised to build a grey water plant, but, y’all won’t believe this… he didn’t. hasn’t even broken ground on the construction.
also, he’s powering grok with generators that have led to ~74% increase in the shittiness of the air quality in memphis from generators’ exhaust pipes.
please do not use grok.
Is it really that much different from the types of people who would believe the first (often bias-confirming) answer they found on Google?
Its just sad that people who know how internet search works for the sake of comfort let someone else decide what they see when searching for stuff.
Stupid people and internet illiterate people exist since the dawn of the internet. Thats not news.
Whats scary is how a new generation of people will lose or not develop basic critical thinking and research skills, and let an algorithm serve up content to them when looking for information.
I’m one of those. I would prefer to google for results, but current day google is perfectly useless even for very obvious searches…
Then use a different search engine… Don’t rely on AI to use the internet.
Google has become shit enough that you may as well.
Yet another tragedy of our time is watching Google ruin the internet with its SEO bullshit and then replace it with AI
MFW someone says “chat gpt it”:
Comment section under every post in the god forsaken place formerly known as Twitter.
What’s this show called again?
It’s from Finding Nemo
Thanksies!
deleted by creator
Lemmy
Absolutely funny response, LOL!!!
I wonder why you removed it…
Oh shit nice I had no idea this subdomain exists, guess I’ll finally be rid of those stupid ai answers in private windows as well!
And here’s my geriatric ass still saying: “look it up on a search engine”
I just say look it up
go back in time and look it up in 2015 before the web wa[ENTER YOUR EMAIL ADDRESS TO SUBSCRIBE AND READ THE REST OF THIS POST]*
archive.is is your friend.
Archive.org could actually build a genuinely useful fully self contained search engine using the real pages and media they have.
“the answer’s on the information superhighway!”
Well that’s because in the years leading up to the release of ChatGPT Google became shittier and shittier in performing the service everyone was trying to use it for and everyone needed to work. Then came out chatbots, which basically gave that functionality back. For most people accuracy didn’t matter because they could finally feel again that going to the internet to learn about something is viable.
Then of course AI makes search algortihms even less useful and all the ethical problems about AI are raised as well.
But in the end I can’t really blame people who now use AI instead of Google or something. Finding something with just a search query was always rightfully ours and greedy companies took it to make more ad revenue. Now they are finally giving back a worse version of the original, but one that’s still far superior to what search engines devolved into by now. It’s no surprise most people love and also need that.
There is at least some merit to the technology for this use-case too though (doesn’t even remotely justify the energy costs though, of course). It’s one of the few things LLMs are genuinely good at since it merely requires text ingestion and to regurgitate what was ingested on some way. As long as it’s paired with proper sources (no clue how ChatGPT does it) for all claimed findings it really can be better. Obviously it’s also “better” since it circumvents all the utterly ridiculous trash we usually have to deal with (pop-ups, ads, dark patterns, registration walls, bad search algorithms etc.) which shouldn’t be used as argument.
Paired with the “Thinking” or “Reflection” feature that simulates some basic thinking process (it even enables these things to count the corrrect amount of ‘b’ in ‘blueberry’, wow!) the results are genuinely good (Disclaimer, I only ever tested that with the free tier of Mistral AI - if you really want to use this stuff at least go to them, they’re bound to EU law). I really get why it becomes so popular, and I’d lie if I said I’d never use it myself. Would still prefer if we weren’t going down this cyberpunk timeline though…
Edit: I’m strictly speaking about using it as a search engine when it does look up websites FOR you and reads through them, not when it makes shit up itself. Only then the stuff I said applies!
Soon with some tweaking the AI answers will serve the shitified shit the SEO driven algorithms shat.
Like: Use Elmer’s TM ©️ glue on your Tombstone Pizza from Kroger to keep your toppings sticky!
So you’ll get ads baked into your wrong answer.
I’ve quite solidly taken to the phrase “Just Web search it” and gently replying “yes, I’ll Web search it” when prompted to “google it”.
The worst part of it is that the person who told you this spelled it wrong.
Don’t worry, you’re not old, they’re just stupid.
“you are simply… Bad product”
My wife’s cousin is doing chat gpt queries always before making decisions. She is not stupid but not bright either. I keep telling her it’s a algorithm and not AI but all I get a blank stare.
The worst to me is still “Oh, I ask him stuff all the time.”
Can we not gender the advanced auto complete, please…
I think if someone told me that they’d just get blocked.
I use gpt every day of my life, professionally and personally. I won’t ask an LLM anything I don’t already know the answer to. I’ve seen them hallucinate way too much crap. With the Internet at our fingertips, there’s no excuse not to do a little of your own homework and proof what comes out of your LLMs.
Oh, and remove those em dashes. Everyone knows you don’t write that LinkedIn post full of em dashes, bro.
es and the whole “I’ll tell you your idea’s advantages without disagreeing” behavior are actually still just signs to alert the average person that the text was AI-generated. Sure, you can still choose to ask the LLM to give the text a style and escape human eyes, but I’m assuming these behaviors still exist genuinely for safety reasons.I won’t ask an LLM anything I don’t already know the answer to.
Why are you asking if you already know the answer?
Just go to your local library and look for books on the subject. 🤷♂️
I like to start asking around until someone knows about it and then interview them.
What about information that is less than a year old?
Kagi that shit my guy
