Arum Natzorkhang (COMMS OPEN)
Region: US
Saturday 22 June 2024 00:26:50 GMT
113845
15427
241
178
Music
Download
Comments
MacDoc-101 :
You can quite literally tell chatGPT “no you’re wrong” and it will change its answer to whatever you want pretty much
2024-06-22 12:13:24
2165
Dramius :
The fact is ChatGPT and OpenAI need to do a better job of letting people know that these AI hallucinate 90%+ of everything they produce.
2024-06-22 01:20:32
1170
ada :
also not all cognates have the same meaning
2024-08-25 22:56:36
0
Kendrick :
Hold up, how do you say “miyaw”?
2024-06-22 00:52:28
67
Dolemaxxer :
AI has a LOT of potential, but in its current state, it has no place in serious academics. In decades? I’m sure it will be invaluable, but it’s nowhere near advanced enough now.
2024-06-22 00:34:20
397
Jesse Lucas :
google and chat gpt benefit from lying to people, because giving false info that looks real makes them look more knowledgeable/ better than saying "i don't know"
2024-06-22 06:46:14
186
Short one :
AI creates "pictures" it doesn't create actual inteligent content it monkeys actions. Good example is Ai art, it's based on visuals, but not on theory behind them
2024-06-22 10:38:12
143
Slippin' Jammy (tadpoled) :
people think AI is magic rather than just next token prediction
2024-06-22 13:37:00
74
notmyrealname82 :
yes, artificial hallucinations are really the Bain of of modern ai. less proof test data less likely to be correct
2024-06-22 01:05:48
37
Ovrwrite :
I work in software and with AI. It's alarming how often I have to tell people that chatGPT is not a magical truth box
2024-06-24 15:08:27
15
pickledxu :
There are actually many serious advocates of AI in academia who understand the potential and limitations of AI. My U recently hosted a local AI conference and the ideas exchange were fascinating
2024-06-22 03:07:08
10
Sam Strosity 🍉 :
Google translate is notoriously terrible if you actually speak the language it is translating to.
2024-06-22 00:54:30
21
Jack_Man99 :
Lol welcome to the world my guy !
2024-06-26 17:06:48
7
Aaron :
Autoregressive token prediction can be really useful and I've found GPT4 to be useful in doing research as something I can bounce ideas off of - but it just increases productivity, not capability.
2024-06-23 09:51:11
0
ayaacount :
In human scholars we trust 👏👏👏
2024-06-22 13:23:05
7
Vaughn :
Google Translate is a trickster spirit. 😂
2024-06-24 15:41:16
2
simoaneorsimon :
that's not even the name for cat it's something like Meu cause it sounds like meow
2024-06-23 01:56:50
7
aGGeRReS :
Google translate was using some sort of Word2Vec or embeddings for many years. As it was often failin gto translate some obvious things.
2024-06-22 08:38:50
0
Helmore Flames :
Omg it's kind of a given how limited google translate or even deepl can be.
2024-06-22 05:33:47
2
Rob :
Chstgpt is only good for general basic info more specialized info it’s better if you already have a basis in it but then that kinda defeats its purpose
2024-06-30 18:06:32
5
Debbie Day :
That's why I only use ChatGPT for writers block
2024-06-22 16:55:19
3
OptimalityTheo :
Hittite Google translate is such a cursed concept
2024-06-22 15:10:08
3
zkfilmchannel :
I use chatgpt to reverse search books on very specific topics and it HAS to be double checked because it hallucinates a lot of information.
2024-07-01 23:38:59
2
Calded :
no body gets that chat gpt if a tool to help you do things, not something that does things for you. it wont find out stuff for you accuratly cuz that's not what it was designed for.
2024-07-03 15:23:23
4
jasongarlido :
Your last line is important. It can only know what it has been trained. It is limited in our own knowledge
2024-06-22 13:10:49
6
To see more videos from user @arumnatzorkhang, please go to the Tikwm
homepage.
