When I tried it in the past, I kinda didn't take it seriously because everything was confined to its instance, but now, there's full-featured global search and proper federation everywhere? Wow, I thought I heard there were some technical obstacles making it very unlikely, but now it's just there and works great! I asked ChatGPT and it says this feature was added 5 years ago! Really? I'm not sure how I didn't notice this sooner. Was it really there for so long? With flairs showing original instance where video comes from and everything?

you are viewing a single comment's thread
view the rest of the comments

[–] spankmonkey@lemmy.world 10 points 4 days ago* (last edited 4 days ago) (4 children)

google: I checked the listing of news sites to find information about a world event directly from professionals who double check their sources

chatGPT: I asked my hairstylist their uninformed opinion on a world event based on overheard conversations

I mean a moron could find the wrong information from google and your hairstylist could get lucky and be right, but odds are one source provides the opportunity for reliable results and the other is random and has a massive shit ton of downsides.

[–] Flax_vert@feddit.uk 1 points 4 days ago

Google results are like:

Is peertube compatible with the fediverse?

ADVERT

Introduction: A lot of people wonder if peertube works with other peertube instances....

ADVERT

What is peertube? Peertube was set up in 1989 by john Peer...

Pop-up: do you like our publication? Give us your email address.

ADVERT

Why you might want to set up peertube: peertube is a decentralised way....

ADVERT

Please support us! From £30 a month you can help us to write more.

Wat is the fediverse? The fediverse is a technology...

ADVERT

Articles you may also like:

How to install Microsoft Teams
How to rent servers from Amazon
How to enable all data collection on Google

ADVERT

So can peertube instances talk to each other?

ADVERT

the answer is yes.

ADVERT

In conclusion, peertube is very...

Comments (169)

John Smith wrote at 12:28 on Friday

Peertube is actually developed by a communist who turned my daughter gay. Boycott!!!

[–] kudra@sh.itjust.works -1 points 4 days ago

What if your hairstylist is on the Fediverse, avoids mainstream social media, and spends a lot of their spare time reading scientific papers?

[–] LesserAbe@lemmy.world -3 points 4 days ago (2 children)

Lots of legitimate concerns and issues with AI, but if you're going to criticize someone saying they used it you should at least understand how it works so your criticism is applicable.

It is useful. Chatgpt performs web searches, then summarizes the results in a way customized to what you asked it. It skips the step where you have to sift through a bunch of results and determine "is this what I was looking for?" and "how does this apply to my specific context?"

Of course it can and does still get things wrong. It's crazy to market it as a new electronic god. But it's not random, and it's right the majority of the time.

[–] Stillwater@sh.itjust.works 10 points 4 days ago (3 children)

It might be wrong more often than you think

https://futurism.com/study-ai-search-wrong

[–] thedruid@lemmy.world 7 points 4 days ago

IS wrong

Ftfy

[–] LesserAbe@lemmy.world 3 points 4 days ago

Besides the other commenter highlighting the specific nature of the linked study, I will say I'm generally doing technical queries where if the answer is wrong, it's apparent because the AI suggestion doesn't work. Think "how do I change this setting" or "what's wrong with the syntax in this line of code". If I try the AI's advice and it doesn't work, then I ask again or try something else.

I would be more concerned about subjects where I don't have any domain knowledge whatsoever, and not working on a specific application of knowledge, because then it could be a long while before I realize the response was wrong.

[–] hisao@ani.social 1 points 4 days ago (1 children)

In this study they asked to replicate 1:1 headline publisher and date. So for example if AI rephrased headline as something synonymous it would be considered at least partially incorrect. Summarization doesn't require accurate citation, so it needs a separate study.

[–] Stillwater@sh.itjust.works 8 points 4 days ago* (last edited 4 days ago)

OK but google (or ask your AI?) about AI accuracy. This isn't the only source saying theres a problem with the answers.

[–] grue@lemmy.world 10 points 4 days ago (1 children)

It skips the step where you have to sift through a bunch of results and determine "is this what I was looking for?" and "how does this apply to my specific context?"

Right: it skips the part where human intelligence and critical thinking is applied. Do you not understand how that's a fucking problem‽

[–] LesserAbe@lemmy.world 0 points 4 days ago

Could you try to understand what I'm saying instead of jumping down my throat?

If I want to turn off a certain type of notification in a program I'm using, I don't need to sift through three forum threads to learn how to do that. I'm fine taking the AI route and don't think I've lost my humanity.

[–] Eyekaytee@aussie.zone -4 points 4 days ago* (last edited 4 days ago)

chatGPT: I asked my hairstylist their uninformed opinion on a world event based on overheard conversations

LLM's have been able to search the web for a few years now

The main one outside of ChatGPT is https://www.perplexity.ai/

You can also look at hosting your version with Local LLM's:

https://olares.medium.com/building-a-local-perplexity-alternative-with-perplexica-ollama-and-searxng-71602523e256