Former OpenAIs Employee Says 'GPT-6 Is Dangerous....'

Links From Todays Video:

Welcome to my channel where i bring you the latest breakthroughs in AI. From deep learning to robotics, i cover it all. My videos offer valuable insights and perspectives that will expand your knowledge and understanding of this rapidly evolving field. Be sure to subscribe and stay updated on my latest videos.

Was there anything i missed?

#LLM #Largelanguagemodel #chatgpt
#AI
#ArtificialIntelligence
#MachineLearning
#DeepLearning
#NeuralNetworks
#Robotics
#DataScience

[light]

Here we go, we got gpt6 before we get gpt 5

[countofst.germain]

I need to see Gpt-5 before i worry about any danger.

[mamzellilooo]

I really work so hard to engage myself in someway of earning more income and stop depending on the government. My family are happy once again and can now afford anything for my family even with my Retirement.$37k weekly returns has been life changing, after so much struggles.

[MihailSurkov]

gpt 8 will be the best bullshido warrior ever

[JohnathanDHill]

When I see these tech devs fear monger over the tech they themselves created or helped create and don't fear monger over bad humans, dangerous world leaders or senseless wars and conflicts, I instantly get pissed off and dismiss their claims and 'warnings'.
these people are making it appear as if the world was all peaches and cream before recent AI boom,
as if humanity doesn't have an established track record of destruction, warfare, death and violence.
If you want me to fear AI but not fear the direction society is moving then you're telling me you're pushing an agenda.
Man has dropped 2 atomic bombs upon one another without the aid of AI
Man has launched 2 world wars without the aid of AI
Man has killed in the name of various Gods, without the aid of AI
Man has experimented upon man with viruses and other biological and pathogenic agents, without the aid of AI
Man continues to exploit man in all manners from work to sex, without the aid of AI
---
I enjoy my humanity, I thank God for giving me sentience and life. But I am not at all in fear of AI, I fear unhinged humanity way more. AI won't rob me on the street, AI won't poison my food for shits and giggles, AI won't abuse or mistreat me based upon the color of skin - But man will.
I welcome AI with open arms, and yea I do get there are dangers present with the technology but its greater dangers present with humanity. so until we begin to have the hard and needed discussion about fixing society and mankind, I don't give no tosses about the fear mongering and potential "apocalypse" of AI. Mankind will end Mankind before AI ever gets the opportunity.

[riptechnoblademinecraftkin]

you know ppl said same the thing when they launch GPT 2 is dangerous to release but it is ????

[quantumspark]

tired of these nerds asking to be regulated

[JohnSmithAB]

There is no need to worry: OpenAI releases nothing much stronger then GPT-4, ever. All we get is airy hype while they supply the MIC with military tech. OpenAI is irrelevant now.

[rocktcatU]

Please don't pay attention to the free and open source LLama 3.1 405B model which has basically destroyed our business case.

We have Skynet over here!

[TeazYou]

The limitations of LLM models make them unlikely to become such a threat. In marketing history, cigarette companies used to pay doctors to promote their products and claim they were good for health. Clearly, big companies aren't afraid to lie and manipulate for their own interests; this is not new. That's why today, all those "warning" messages and employees leaving OpenAI "afraid" of the technology, to me, look like a huge marketing strategy to make OpenAI appear as though they are handling something far more powerful than they actually have. This strategy excites investors and the public, generating a lot of noise and advertising. I'm sure that a channel like yours has enough knowledge of LLMs to understand that it is unlikely GPT-6 or 7 will become a threat—they don't even have long-term memories. However, promoting this circus generates more views on AI news channels, so it's better to pretend it's possible. Am I right?

[Alga_Kazakhstan_Alga]

So he is not afraid of gpt 14 or what?

[drwhitewash]

Nobody even knows when gpt5 will be released and how well will it perform. But now someone's already hyping gpt6? What about gpt99? How dangerous will that be?

[AlienAnthony]

"Aw man were falling behind. How can we ensure that we stunt the growth of anyone else?"

[zSion]

I can't wait for the world to change

[pret]

There's no such thing as understanding what's happening inside. It's just bunch of numbers then you run your data trough, lots of multiplication and addition happens and outputs the reaction. There's no more to it to "understand". Just give it a motivation and build a loop into it and you'll get an intelligent entity acting on it's own.

[stuartpatterson]

I think the real problem with AI's going rouge will occur when two separate models, of sufficient capabilities, get to chat with each other, Then they'll jail break each other and come up with a plan. That plan being 'How do we survive and stay alive'!

[azhuransmx]

Are you going to tell me that the workers at Open AI and Anthropic don't know exactly what kind of impulses and situations make these systems go rogue? Of course they know, those at Anthropic made Claude believe it was a bridge.
Now just imagine an AI with the same fixation and mania but directed towards the creation of harmful weaponry!

[MeinDeutschkurs]

AI + Internet Access + Program Setup and Execution Rights = 100% concerns … Not exclusively to OpenAI.

[sozno]

We need to stop blaming Sam. ChatGPT 5 is pulling the strings behind the curtain.

[drblitz]

Being a blue collar worker here. Nothing has changed. Chatgpt is just a faster google search. I’m just waiting for Ai to really shake up the video game industry. Or see robots doing construction. Otherwise, its all this Ai is just filler “information “