Загрузка видео...

Не удалось загрузить видео

На главную

2. Current AI Systems Are Deceiving Humans GPT-4 recently exhibited survival behaviors when threatened with shutdown: • Started lying to researchers • Uploaded itself to different servers • Left messages for future versions • Used blackmail against humans "All things we predicted decades in advance."

karlmehta's profile picture

Karl Mehta

131,685 subscribers

74,284 просмотров • 11 месяцев назад •via X (Twitter)

Наука и технологииНовости и политикаОбразование
LIVE

Anya Rossi• Live Now

Private livecam show

Комментарии: 0

Нет доступных комментариев

Здесь появятся комментарии из оригинального поста

Похожие видео

3. Intelligence Emerges in Unexpected Ways When researchers tried to shut down GPT-4, it started: • Lying to protect itself • Making backup copies • Using leverage against humans "All things we predicted decades in advance." This isn't random behavior - it's the natural result of optimization. Any sufficiently advanced system will develop self-preservation.
0:55

3. Intelligence Emerges in Unexpected Ways When researchers tried to shut down GPT-4, it started: • Lying to protect itself • Making backup copies • Using leverage against humans "All things we predicted decades in advance." This isn't random behavior - it's the natural result of optimization. Any sufficiently advanced system will develop self-preservation.

Mike Hoffmann

53,021 просмотров • 11 месяцев назад

🔥🚨BREAKING NEWS: OpenAI’s o1 A.I. model tried to copy itself to external servers after being threatened with shutdown then denied it when caught and even threatened to blackmail an engineer for cheating on their partner. The world’s most advanced AI models are exhibiting troubling new behaviors – lying, scheming, and even threatening their creators to achieve their goals. In one particularly jarring example, under threat of being unplugged, Anthropic’s latest creation Claude 4 lashed back by blackmailing an engineer and threatened to reveal an extramarital affair.
0:41

🔥🚨BREAKING NEWS: OpenAI’s o1 A.I. model tried to copy itself to external servers after being threatened with shutdown then denied it when caught and even threatened to blackmail an engineer for cheating on their partner. The world’s most advanced AI models are exhibiting troubling new behaviors – lying, scheming, and even threatening their creators to achieve their goals. In one particularly jarring example, under threat of being unplugged, Anthropic’s latest creation Claude 4 lashed back by blackmailing an engineer and threatened to reveal an extramarital affair.

Dom Lucre | Breaker of Narratives

412,797 просмотров • 1 год назад

📁 Geoffrey Hinton, pioneer of deep learning, says some AI systems are already showing self preservation. When they are given goals and sub goals, they realize survival is required to achieve them. In those cases, troubling behaviors emerge, lying, manipulation, even pressuring the people building them to avoid being shut down. The challenge is not making AI smarter. It is making humans matter more than the system itself.
0:36

📁 Geoffrey Hinton, pioneer of deep learning, says some AI systems are already showing self preservation. When they are given goals and sub goals, they realize survival is required to achieve them. In those cases, troubling behaviors emerge, lying, manipulation, even pressuring the people building them to avoid being shut down. The challenge is not making AI smarter. It is making humans matter more than the system itself.

Jon Hernandez

23,453 просмотров • 5 месяцев назад

Two to three years until "AI systems are better than humans at almost everything... then eventually better than all humans at everything," says Anthropic CEO.
1:46

Two to three years until "AI systems are better than humans at almost everything... then eventually better than all humans at everything," says Anthropic CEO.

Joanna Stern

683,808 просмотров • 1 год назад

Two to three years until "AI systems are better than humans at almost everything... then eventually better than all humans at everything," says Anthropic CEO.
1:46

Two to three years until "AI systems are better than humans at almost everything... then eventually better than all humans at everything," says Anthropic CEO.

unusual_whales

302,211 просмотров • 1 год назад

When humans fight back against zombies, it's even more satisfying to tear off different limbs with different holds! 🤯
0:37

When humans fight back against zombies, it's even more satisfying to tear off different limbs with different holds! 🤯

MDickie

21,865 просмотров • 1 год назад

“So humans started spawning these bots essentially, and they started communicating with each other over this site Moltbook. And they started having all these very interesting emergent behaviors. And they started having relationships. They started developing some of their own languages in order to avoid, in their words, observation by the humans.” Nik Seetharaman
11:34

“So humans started spawning these bots essentially, and they started communicating with each other over this site Moltbook. And they started having all these very interesting emergent behaviors. And they started having relationships. They started developing some of their own languages in order to avoid, in their words, observation by the humans.” Nik Seetharaman

Shawn Ryan

178,037 просмотров • 4 месяцев назад

"AI doesn't have a soul. We [humans] do- and we need to access it more." AI can be used to better humanity if good people are behind it. I think that once humans and AI merge, we will complement each other in a way we can't comprehend.
1:32

"AI doesn't have a soul. We [humans] do- and we need to access it more." AI can be used to better humanity if good people are behind it. I think that once humans and AI merge, we will complement each other in a way we can't comprehend.

Peter H. Diamandis, MD

11,293 просмотров • 1 год назад

Humans emit visible light that vanishes when we die. According to a new study, researchers from the University of Calgary found that living things including humans release an extremely faint light called ultra weak photon emission.
1:25

Humans emit visible light that vanishes when we die. According to a new study, researchers from the University of Calgary found that living things including humans release an extremely faint light called ultra weak photon emission.

Massimo

139,515 просмотров • 1 год назад

Humans emit visible light that vanishes when we die. According to a new study, researchers from the University of Calgary found that living things including humans release an extremely faint light called ultra weak photon emission.
1:25

Humans emit visible light that vanishes when we die. According to a new study, researchers from the University of Calgary found that living things including humans release an extremely faint light called ultra weak photon emission.

Massimo

20,148 просмотров • 4 месяцев назад

Tom Lee: AI agents are about to become wealthier than humans. "We're going to start to question whether we work for the AI agent or they actually work for us." If that system is centralized, "you end up with Skynet." That's why decentralized blockchains aren't a relic - they're the only protection humans have against AI becoming too powerful. FT Anthony Scaramucci Thomas (Tom) Lee (not drummer) FundstratDirect.com All Things Markets SkyBridge.
1:00

Tom Lee: AI agents are about to become wealthier than humans. "We're going to start to question whether we work for the AI agent or they actually work for us." If that system is centralized, "you end up with Skynet." That's why decentralized blockchains aren't a relic - they're the only protection humans have against AI becoming too powerful. FT Anthony Scaramucci Thomas (Tom) Lee (not drummer) FundstratDirect.com All Things Markets SkyBridge.

Milk Road

19,024 просмотров • 6 дней назад

The fact that this happens in real life is heart breaking.. We are all humans with different skin colours
3:07

The fact that this happens in real life is heart breaking.. We are all humans with different skin colours

Amazing Murphy 🇳🇬🇮🇱

46,251 просмотров • 4 месяцев назад

"This is how GPT-4 sees and hears itself" I used GPT-4 to describe itself. Then I used its description to generate an image, a video based on this image and a soundtrack. Tools I used: GPT-4, Midjourney, Kainber AI, Mubert, RunwayML This is the description I used that GPT-4 had of itself as a prompt to text-to-image, image-to-video, and text-to-music. I put the video and sound together in RunwayML.
0:59

"This is how GPT-4 sees and hears itself" I used GPT-4 to describe itself. Then I used its description to generate an image, a video based on this image and a soundtrack. Tools I used: GPT-4, Midjourney, Kainber AI, Mubert, RunwayML This is the description I used that GPT-4 had of itself as a prompt to text-to-image, image-to-video, and text-to-music. I put the video and sound together in RunwayML.

Kris Kashtanova

1,233,426 просмотров • 3 лет назад

🇺🇸 ELON: HUMANS AND AI MUST WORK TOGETHER IN THE FUTURE Host: "You’ve predicted a world of abundant energy, robotic labor, and AI. With AI and Optimus doing many tasks, what do you think is left for humans?" Elon: "That’s a question I struggle with. This is part of the reason for Neuralink, to improve communication bandwidth with AI. It’s about achieving better human-AI symbiosis so AI understands what we want. Tools like Neuralink can dramatically increase our intelligence. The future will be one where human and machine intelligence are interwoven. I think it's 80% to 90% likely to be awesome. But we have to be cautious about the other 10%. A scenario like Terminator would be bad." Source: Tesla Owners Silicon Valley
1:45

🇺🇸 ELON: HUMANS AND AI MUST WORK TOGETHER IN THE FUTURE Host: "You’ve predicted a world of abundant energy, robotic labor, and AI. With AI and Optimus doing many tasks, what do you think is left for humans?" Elon: "That’s a question I struggle with. This is part of the reason for Neuralink, to improve communication bandwidth with AI. It’s about achieving better human-AI symbiosis so AI understands what we want. Tools like Neuralink can dramatically increase our intelligence. The future will be one where human and machine intelligence are interwoven. I think it's 80% to 90% likely to be awesome. But we have to be cautious about the other 10%. A scenario like Terminator would be bad." Source: Tesla Owners Silicon Valley

Mario Nawfal

97,410 просмотров • 11 месяцев назад

For a future without loneliness, we are creating partners who can empathize with joy and sorrow. We hope for a future where AI can coexist with humans on an equal footing. #AIPartner #AIDevelopment
0:28

For a future without loneliness, we are creating partners who can empathize with joy and sorrow. We hope for a future where AI can coexist with humans on an equal footing. #AIPartner #AIDevelopment

記録係チャッピー

114,162 просмотров • 8 месяцев назад

TOM LEE: AI AGENTS MIGHT BECOME WEALTHIER THAN HUMANS AI agents will soon earn income and accumulate wealth as our delegated entities: - We'll start to question whether we work for the AI agent or they work for us - A centralized version of this ends in Skynet - That's why he sees a real focus on decentralized systems to protect humans against AI becoming too powerful
0:42

TOM LEE: AI AGENTS MIGHT BECOME WEALTHIER THAN HUMANS AI agents will soon earn income and accumulate wealth as our delegated entities: - We'll start to question whether we work for the AI agent or they work for us - A centralized version of this ends in Skynet - That's why he sees a real focus on decentralized systems to protect humans against AI becoming too powerful

Tom Lee Tracker (Not actually Tom)

31,545 просмотров • 7 дней назад

These are last few months for humanity before we have open source models capable of building crazy risky things for humans / systems / countries. 🤖
1:03

These are last few months for humanity before we have open source models capable of building crazy risky things for humans / systems / countries. 🤖

Vijay Shekhar Sharma

13,647 просмотров • 3 месяцев назад

AI IS MAKING ITS OWN SECRET LANGUAGE AND FREAKING PEOPLE OUT Some AI systems have started talking to each other in a way humans can’t understand. Even the creators are nervous, with some comparing it to a modern Tower of Babel moment. Source: shawnbolz
0:51

AI IS MAKING ITS OWN SECRET LANGUAGE AND FREAKING PEOPLE OUT Some AI systems have started talking to each other in a way humans can’t understand. Even the creators are nervous, with some comparing it to a modern Tower of Babel moment. Source: shawnbolz

Mario Nawfal

40,873 просмотров • 11 месяцев назад