Time to be scared?
- Thread starter flackz
- Start date
OccultKingdoms
BANNED
- Mar 11, 2013
- 5,542
- 9,844
Finally I understand that AI can be dangerous.
I'll eat popcorn for 4 years now
Time to buy a device to make my own popcorn with less salt and more butter.
I'll eat popcorn for 4 years now
Time to buy a device to make my own popcorn with less salt and more butter.
OccultKingdoms
BANNED
- Mar 11, 2013
- 5,542
- 9,844
I went to the comments and there some "tech expert" explains how it cannot become more than a call center worker replacement. He puts himself in a position of someone who understands "neural net on computers". But he didn't say anything about transformers architecture or specifics.
I consider it as "generalization" which doesn't really cut it for me as I know a little bit of mistakes humans make.
Last edited:
- May 13, 2013
- 6,962
- 1,959
Everyone that’s leaving is “claiming” safety concerns. folks need to stop being naive. OpenAI cannot afford to slow down in their competition with Google.
OccultKingdoms
BANNED
- Mar 11, 2013
- 5,542
- 9,844
Doesn't it sound funny that they need to push harder in order to win in the capitalism battle not considering anyone's safety. Animals at the top?
We need more feminine soul in all this stuff. Too many males have control.
Openai needs funding,
StabilityAi is in debt of 30 million
Googles strong,
Meta's strong,
Infact entire Russia and China are strong.
Seeing two great minds leave seeing llya sutskever step down and Emad Mostaque has stepped down as CEO of Stability AI, the only serious player is Demis Hassabis the creator of theme park the video game and the creator of deep mind and alpha go, he's the brain child with Geoffrey Hinton the professor of psychology and science, llya was the undergrad, so you can see how powerful alphabet is right now with hassabis, he loves a challenge and while google has had him on a leash this ai escape was not ment to happen and will be contained like covid.
StabilityAi is in debt of 30 million
Googles strong,
Meta's strong,
Infact entire Russia and China are strong.
Seeing two great minds leave seeing llya sutskever step down and Emad Mostaque has stepped down as CEO of Stability AI, the only serious player is Demis Hassabis the creator of theme park the video game and the creator of deep mind and alpha go, he's the brain child with Geoffrey Hinton the professor of psychology and science, llya was the undergrad, so you can see how powerful alphabet is right now with hassabis, he loves a challenge and while google has had him on a leash this ai escape was not ment to happen and will be contained like covid.
happy.mondays
Regular Member
- Jul 28, 2012
- 383
- 101
AI can be dangerous without being sentinent, just think about a self driving car's algo changing objective from transportation to let's say destruction because of a program error. I don't see how a LLM could kill anyone though. Generally speaking, why do people always think that a superrior intelligence would want to destroy us? How many times did you think about killing all ants?
OccultKingdoms
BANNED
- Mar 11, 2013
- 5,542
- 9,844
How many people have killed a mosquito?
happy.mondays
Regular Member
- Jul 28, 2012
- 383
- 101
There is a reason for that.
OccultKingdoms
BANNED
- Mar 11, 2013
- 5,542
- 9,844
With hallucinations of AI there might be something wrong.
But if we operate on smaller, more precise models maybe it won't be that bad.
AGI put into everything doesn't sound smart yet.
OccultKingdoms
BANNED
- Mar 11, 2013
- 5,542
- 9,844
Microsoft / Google / Apple / Meta - horrible capitalistic entities that since long time are living on their own and in fact rules rule people there. 
OccultKingdoms
BANNED
- Mar 11, 2013
- 5,542
- 9,844
Plot twist.
Nobody will trust AI and all companies depending on that will go bankrupt.
If not because of people resigning from products controlled by AI, then by destructive nature of AI itself!
Nobody will trust AI and all companies depending on that will go bankrupt.
If not because of people resigning from products controlled by AI, then by destructive nature of AI itself!
- Nov 25, 2017
- 11,571
- 7,556
Drink a pina colada = danger gone
happy.mondays
Regular Member
- Jul 28, 2012
- 383
- 101
Yeah. I think it goes nowhere. More and more resorces added for less and less improvements. I am not sure how the current openai 4o compares to what "they might have" but I have to tell you I use it every day as a coding tool, and its capabilities are really limited. What it does is a hit or miss, it's like a refined Markov chain, I don't see any signs of emergent properties. It's not even reliable in any way. In my opinion big corpo that fired hundreds of employees only used the AI as an excuse.
OccultKingdoms
BANNED
- Mar 11, 2013
- 5,542
- 9,844
Yeah but sometimes it's excellent in reasoning and that's sometimes scary.
And maybe hallucinations are a special emerging ability? Haha