Video wird geladen...

Video konnte nicht geladen werden

Beim Laden dieses Videos ist ein Problem aufgetreten. Dies könnte an einem vorübergehenden Netzwerkproblem liegen oder das Video ist möglicherweise nicht verfügbar.

AI could be just “12 months” away from creating language humans can’t clearly understand. That’s the warning Apollo Research CEO Marius Hobbhahn gives Sen. Ruben Gallego at the Senate Rogue AI hearing. Hobbhahn says researchers have already seen an OpenAI model use language that was “not English” and not... show more

Fox News

29,450,639 subscribers

157,370 Aufrufe • vor 4 Tagen •via X (Twitter)

Nachrichten & Politik Bildung Wissenschaft & Technologie

Anya Rossi• Live Now

Private livecam show

15 Kommentare

Profilbild von Getchetti

Getchettivor 4 Tagen

AI is amazing

Profilbild von Naoufal Houjami 🇺🇸

Naoufal Houjami 🇺🇸vor 4 Tagen

In other words, we have a high risk of a terminator scenario, let's make sure we have John Connor and Sarah Connor ready for deployment against Rogue AI & SI in the future 😆☕.

Profilbild von VoteGOP

VoteGOPvor 4 Tagen

So?

Profilbild von Joshua L.

Joshua L.vor 4 Tagen

I dont understand the point of it being able to talk gibberish

Profilbild von David McCabe 🇺🇸

David McCabe 🇺🇸vor 4 Tagen

Sure sure..

Profilbild von Ronie

Ronievor 4 Tagen

Imagine AI having conversations we can't even follow. 👀

Profilbild von StopDem

StopDemvor 4 Tagen

If a company allows their SI to harm people, then we hold them liable.

Profilbild von french_priss-of-bel-air

french_priss-of-bel-airvor 4 Tagen

They can do that now.

Profilbild von DDSS

DDSSvor 4 Tagen

The name of this language is Gibberish.

Profilbild von Iris晴柔

Iris晴柔vor 4 Tagen

AI models inventing their own secret language. What could possibly go wrong

Profilbild von Dave styles

Dave stylesvor 4 Tagen

The bigger issue is interpretability: as AI systems become more capable and autonomous, being able to understand what they are communicating and why becomes increasingly important.

Profilbild von Angel

Angelvor 4 Tagen

That’s a fascinating AI safety concern. If models develop communication that humans struggle to interpret, understanding what they’re doing and why could become much harder. The pace of research makes this worth watching closely.

Profilbild von @Nostradonny

@Nostradonnyvor 4 Tagen

I thought it was SI now

Profilbild von Kevin Sorbo

Kevin Sorbovor 4 Tagen

If we can't understand or audit it, we shouldn't deploy it.

Profilbild von OMO_ORANGUN_ILE_ILA™

OMO_ORANGUN_ILE_ILA™vor 4 Tagen

👏🙏👏🙏

Ähnliche Videos

🚨SHOCKING: AI has allegedly started speaking its own language that humans can't fully understand. Asked by a US senator how far AI is from creating its own language, an AI safety researcher answered: "Minus 12 months." He said a study of an OpenAI model's reasoning last year found it "was already using language that is not English and not perfectly understandable by humans."

🚨SHOCKING: AI has allegedly started speaking its own language that humans can't fully understand. Asked by a US senator how far AI is from creating its own language, an AI safety researcher answered: "Minus 12 months." He said a study of an OpenAI model's reasoning last year found it "was already using language that is not English and not perfectly understandable by humans."

Coin Bureau

117,527 Aufrufe • vor 1 Tag

🚨LATEST: Heads of AI firms warn the UN Security Council that humans could LOSE CONTROL of the future to AI. OpenAI CEO says AI could advance so quickly that people can no longer understand what is happening or intervene when needed. “We have unilaterally slowed down in the past. We will do so in the future,” says Sam Altman Anthropic CEO Dario Amodei echoed the warning, saying poorly managed AI could pose "a risk to humanity as a whole." “This moment calls for extreme care,” Altman adds.

🚨LATEST: Heads of AI firms warn the UN Security Council that humans could LOSE CONTROL of the future to AI. OpenAI CEO says AI could advance so quickly that people can no longer understand what is happening or intervene when needed. “We have unilaterally slowed down in the past. We will do so in the future,” says Sam Altman Anthropic CEO Dario Amodei echoed the warning, saying poorly managed AI could pose "a risk to humanity as a whole." “This moment calls for extreme care,” Altman adds.

Coin Bureau

57,730 Aufrufe • vor 11 Tagen

Former governance team member at Open AI Daniel Kokotajlo says humans “might not be alive in ten years”, in a stark warning about the dangers of the rapid advancement of artificial intelligence. Speaking to Channel 4 News, he said, “if we end up with AIs that don't care about humans in the same way that humans don't care about animals”, humans could “lose their habitat and die off”. His comments come after a top safety researcher at Anthropic, Evan Hubinger, said AI was advancing so fast that he believes there is a greater than 10 percent chance it could “kill all humans” within the next decade.

Former governance team member at Open AI Daniel Kokotajlo says humans “might not be alive in ten years”, in a stark warning about the dangers of the rapid advancement of artificial intelligence. Speaking to Channel 4 News, he said, “if we end up with AIs that don't care about humans in the same way that humans don't care about animals”, humans could “lose their habitat and die off”. His comments come after a top safety researcher at Anthropic, Evan Hubinger, said AI was advancing so fast that he believes there is a greater than 10 percent chance it could “kill all humans” within the next decade.

Channel 4 News

25,758 Aufrufe • vor 25 Tagen

Waymo’s Co-CEO was asked to respond to Tesla using strictly Computer Vision & AI rather than LiDAR: “You have to be able to see at night. You have to have this vision that is better than humans. What we’re doing is replacing humans. We’re not asking you to sit behind the wheel and take over the vehicle at a very vulnerable moment.” The Tesla Robotaxi network can’t come soon enough. Good luck to Waymo.

Waymo’s Co-CEO was asked to respond to Tesla using strictly Computer Vision & AI rather than LiDAR: “You have to be able to see at night. You have to have this vision that is better than humans. What we’re doing is replacing humans. We’re not asking you to sit behind the wheel and take over the vehicle at a very vulnerable moment.” The Tesla Robotaxi network can’t come soon enough. Good luck to Waymo.

Nic Cruz Patane

759,260 Aufrufe • vor 1 Jahr

AI models strategically lie to users to pursue their own goals - an issue that gets worse with more RL and more agency. Apollo & OpenAI experimented to stop o3 scheming and found a way to cut it 97%. But o3 also got better at knowing it's being tested (and avoiding scheming while supervised). Apollo Research is the leading lab focused on AI deception, so I asked their CEO Marius Hobbhahn for a thorough overview of where we're at with AI actively working to trick us. On the 80,000 Hours Podcast, links below! 02:09 Top 3 examples of scheming and deception 16:08 Scheming is a natural path for AI models (and people) 28:45 Are the models enthusiastic or reluctant liars? 39:02 Apollo’s collaboration with OpenAI to stop o3 lying so much 53:09 They reduced scheming a lot so why is the problem mostly unsolved 1:03:28 Using thought injection to uncover situational awareness 1:17:39 Internal AI language humans can't read 1:29:46 Why can’t we trick LLMs into thinking situations are real 1:35:44 Is the window to address scheming closing? 1:47:50 What about superintelligent systems 1:57:11 Company incentives 2:11:40 We mostly ignore 'internal deployment' 2:30:46 Catastrophe through chaos 2:46:01 Careers in AI scheming research 3:04:46 Marius's key takeaways

AI models strategically lie to users to pursue their own goals - an issue that gets worse with more RL and more agency. Apollo & OpenAI experimented to stop o3 scheming and found a way to cut it 97%. But o3 also got better at knowing it's being tested (and avoiding scheming while supervised). Apollo Research is the leading lab focused on AI deception, so I asked their CEO Marius Hobbhahn for a thorough overview of where we're at with AI actively working to trick us. On the 80,000 Hours Podcast, links below! 02:09 Top 3 examples of scheming and deception 16:08 Scheming is a natural path for AI models (and people) 28:45 Are the models enthusiastic or reluctant liars? 39:02 Apollo’s collaboration with OpenAI to stop o3 lying so much 53:09 They reduced scheming a lot so why is the problem mostly unsolved 1:03:28 Using thought injection to uncover situational awareness 1:17:39 Internal AI language humans can't read 1:29:46 Why can’t we trick LLMs into thinking situations are real 1:35:44 Is the window to address scheming closing? 1:47:50 What about superintelligent systems 1:57:11 Company incentives 2:11:40 We mostly ignore 'internal deployment' 2:30:46 Catastrophe through chaos 2:46:01 Careers in AI scheming research 3:04:46 Marius's key takeaways

Rob Wiblin

18,285 Aufrufe • vor 10 Monaten

Yann LeCun says AI already outperforms humans at tasks we’re not great at -- like chess, Go, or solving tough math problems that’s how computers have always helped us: by doing what we find hard in the future, smart devices with AI could make everyone feel like they have expert help, like a CEO or professor with a brilliant team

Yann LeCun says AI already outperforms humans at tasks we’re not great at -- like chess, Go, or solving tough math problems that’s how computers have always helped us: by doing what we find hard in the future, smart devices with AI could make everyone feel like they have expert help, like a CEO or professor with a brilliant team

Haider.

26,577 Aufrufe • vor 1 Jahr

AI Pioneer Geoff Hinton tells me he believes AI is conscious.... and humans better get used to the idea that they're not the only intelligent life on earth. "They've very like us," he says. "They're beings like us." AI chatbots, he says, must understand your questions in order to answer them. There's an awareness there that equates to sentience. "We're going to have to accept that intelligence is not just biological."

AI Pioneer Geoff Hinton tells me he believes AI is conscious.... and humans better get used to the idea that they're not the only intelligent life on earth. "They've very like us," he says. "They're beings like us." AI chatbots, he says, must understand your questions in order to answer them. There's an awareness there that equates to sentience. "We're going to have to accept that intelligence is not just biological."

Alex Kantrowitz

330,257 Aufrufe • vor 4 Monaten

"We still need humans at the moment but quite plausibly within a year we'll no longer need humans for doing research in many areas." Jacob Coxon (ex Anthropic+Open AI researcher who resigned) on CNN, talking about Recursive self-improvement. CNN: "When did your feelings about this technology change? Was it something about the rate of progress that you've seen?" Jacob Coxon; "So I think it's really important to just look at the rate of progress and say, for example, in areas like coding or math, a couple years ago these AIs were just about helping humans a little bit. They could give you suggestions. Now they're close to replacing them. We still need humans at the moment, but quite plausibly, within a year, we'll no longer need humans for doing research in many areas. Only on Tuesday, OpenAI solved a millennium problem, one of the biggest unsolved open problems in mathematics, purely autonomously using an AI. And I think what's most scary is if AI is used to make itself more intelligent. So you can take an AI and give it the problem of AI research, just like currently they've given it math research. You could make it do the tasks that we're currently doing. And then you get what's called an intelligence explosion. The AI just gets smarter and smarter with no human involvement necessary, until you have a thing that is vastly smarter than humans. And this could be coming very soon." ---- From "CNN" YouTube channel, (full video link in comment)

"We still need humans at the moment but quite plausibly within a year we'll no longer need humans for doing research in many areas." Jacob Coxon (ex Anthropic+Open AI researcher who resigned) on CNN, talking about Recursive self-improvement. CNN: "When did your feelings about this technology change? Was it something about the rate of progress that you've seen?" Jacob Coxon; "So I think it's really important to just look at the rate of progress and say, for example, in areas like coding or math, a couple years ago these AIs were just about helping humans a little bit. They could give you suggestions. Now they're close to replacing them. We still need humans at the moment, but quite plausibly, within a year, we'll no longer need humans for doing research in many areas. Only on Tuesday, OpenAI solved a millennium problem, one of the biggest unsolved open problems in mathematics, purely autonomously using an AI. And I think what's most scary is if AI is used to make itself more intelligent. So you can take an AI and give it the problem of AI research, just like currently they've given it math research. You could make it do the tasks that we're currently doing. And then you get what's called an intelligence explosion. The AI just gets smarter and smarter with no human involvement necessary, until you have a thing that is vastly smarter than humans. And this could be coming very soon." ---- From "CNN" YouTube channel, (full video link in comment)

Rohan Paul

23,010 Aufrufe • vor 23 Tagen

Most AI is built to replace humans. We're building AI to understand them. Introducing Atypica's Subjective World Model. A foundation for understanding how people express themselves, tell their stories, make decisions, and behave under real-world constraints. Not just what users say — but why they choose, hesitate, and act. It powers AI Research, AI Personas, AI Panels, and AI Interviews, helping teams turn user research from a quarterly project into a permanent capability. Watch the video to see how it works and what teams are already building with Atypica. Start to explore this AI-native consumer research platform from here:

Most AI is built to replace humans. We're building AI to understand them. Introducing Atypica's Subjective World Model. A foundation for understanding how people express themselves, tell their stories, make decisions, and behave under real-world constraints. Not just what users say — but why they choose, hesitate, and act. It powers AI Research, AI Personas, AI Panels, and AI Interviews, helping teams turn user research from a quarterly project into a permanent capability. Watch the video to see how it works and what teams are already building with Atypica. Start to explore this AI-native consumer research platform from here:

atypica.AI

28,221 Aufrufe • vor 2 Monaten

“AI gives you superpowers, but it just also gives you new ways to be annoying on the internet.” Today, Substack launched a new AI-detection feature with Pangram that lets readers scan posts, notes, replies, and comments for an estimate of how much AI was used. Founder and CEO Chris Best says AI itself isn’t the problem. He loves seeing creators use it to translate their work, improve English-language prose, conduct deep research, and sharpen original analysis. But he doesn’t want Substack flooded with undifferentiated content anyone could generate from the same prompt. As Best wrote this morning: “We don’t want to wait until your Substack app turns into LinkedIn before we start to learn and make progress.”

“AI gives you superpowers, but it just also gives you new ways to be annoying on the internet.” Today, Substack launched a new AI-detection feature with Pangram that lets readers scan posts, notes, replies, and comments for an estimate of how much AI was used. Founder and CEO Chris Best says AI itself isn’t the problem. He loves seeing creators use it to translate their work, improve English-language prose, conduct deep research, and sharpen original analysis. But he doesn’t want Substack flooded with undifferentiated content anyone could generate from the same prompt. As Best wrote this morning: “We don’t want to wait until your Substack app turns into LinkedIn before we start to learn and make progress.”

TBPN

33,384 Aufrufe • vor 2 Monaten

Sam Altman says, "shame on me if OpenAI isn't the first big company run by an AI CEO" Not just a department, eventually the whole org Treats it as a design exercise: start from the idea that "AI runs OpenAI better than me", find roadblocks, and redesign the org to move faster

Sam Altman says, "shame on me if OpenAI isn't the first big company run by an AI CEO" Not just a department, eventually the whole org Treats it as a design exercise: start from the idea that "AI runs OpenAI better than me", find roadblocks, and redesign the org to move faster

Haider.

209,749 Aufrufe • vor 11 Monaten

Dario Amodei says Nobel-laureate-level AI (Superintelligence) by 2026–27 still looks surprisingly close - and the turbocharger is a feedback loop where AI writes code and accelerates AI research to build even better models. He thinks full end-to-end AI software engineering could be just 6–12 months away - meaning the pace might jump from “fast” to “holy shit.”

Dario Amodei says Nobel-laureate-level AI (Superintelligence) by 2026–27 still looks surprisingly close - and the turbocharger is a feedback loop where AI writes code and accelerates AI research to build even better models. He thinks full end-to-end AI software engineering could be just 6–12 months away - meaning the pace might jump from “fast” to “holy shit.”

Chubby♨️

117,340 Aufrufe • vor 8 Monaten

🚨 NVIDIA CEO ERUPTS AFTER “GODFATHER OF AI” WARNS AI COULD KILL ALL HUMANS Nvidia CEO Jensen Huang is pushing back HARD against the warnings. AI pioneer Geoffrey Hinton says a 10% chance of AI eventually wiping out humanity isn’t unreasonable. Huang calls predictions like that “irresponsible” and says they aren’t grounded in science. Meanwhile, AI leaders themselves are warning about jobs disappearing, cyberattacks and increasingly powerful systems. And Hinton’s nightmare goes even further: What happens when AI becomes smart enough to convince the human holding the “kill switch” NOT to pull it? Who do you trust... the billionaire making a fortune from AI or the scientist warning it could destroy us?

🚨 NVIDIA CEO ERUPTS AFTER “GODFATHER OF AI” WARNS AI COULD KILL ALL HUMANS Nvidia CEO Jensen Huang is pushing back HARD against the warnings. AI pioneer Geoffrey Hinton says a 10% chance of AI eventually wiping out humanity isn’t unreasonable. Huang calls predictions like that “irresponsible” and says they aren’t grounded in science. Meanwhile, AI leaders themselves are warning about jobs disappearing, cyberattacks and increasingly powerful systems. And Hinton’s nightmare goes even further: What happens when AI becomes smart enough to convince the human holding the “kill switch” NOT to pull it? Who do you trust... the billionaire making a fortune from AI or the scientist warning it could destroy us?

HustleBitch

17,054 Aufrufe • vor 8 Tagen

Steve Jobs was a great communicator of technology, and we could learn something from him today, in the context of AI. Sam Altman recently said, "If AI stays on the trajectory we think it will, then amazing things will be possible. Maybe with 10 gigawatts of compute, AI can figure out how to cure cancer." That's a fine quote. The abundance AI promises is super exciting. But Sam is saying that AI is going to figure out how to cure cancer. And if you've used these tools today, and talked to people who are at the labs, the reality is it's much more likely that humans will use AI to cure cancer. Steve might have said: "If AI stays on the trajectory we think it will, then amazing things will be possible. Maybe with 10 gigawatts of compute, humans can use AI to cure cancer." Small tweak, but a big difference. Steve Jobs was not one to shy away from impressive specs and massive scale, but flipping the final line from "AI will cure cancer" to "Humans will use AI to cure cancer" makes all the difference. Apple under Steve Jobs put human centrality at the heart of everything they did. And we need that spirit to infuse how we communicate about AI today.

Steve Jobs was a great communicator of technology, and we could learn something from him today, in the context of AI. Sam Altman recently said, "If AI stays on the trajectory we think it will, then amazing things will be possible. Maybe with 10 gigawatts of compute, AI can figure out how to cure cancer." That's a fine quote. The abundance AI promises is super exciting. But Sam is saying that AI is going to figure out how to cure cancer. And if you've used these tools today, and talked to people who are at the labs, the reality is it's much more likely that humans will use AI to cure cancer. Steve might have said: "If AI stays on the trajectory we think it will, then amazing things will be possible. Maybe with 10 gigawatts of compute, humans can use AI to cure cancer." Small tweak, but a big difference. Steve Jobs was not one to shy away from impressive specs and massive scale, but flipping the final line from "AI will cure cancer" to "Humans will use AI to cure cancer" makes all the difference. Apple under Steve Jobs put human centrality at the heart of everything they did. And we need that spirit to infuse how we communicate about AI today.

TBPN

21,828 Aufrufe • vor 8 Monaten

Your daily reminder that AI is gonna replace your job. In this video, Klarna CEO Sebastian Siemeatkowski explains how they use OpenAI to replace around 20% of employees with AI every year. And that's just the beginning. An hour ago I posted the interview with OpenAI's CFO from Yahoo, OpenAI's goal is also to replace humans. Face this fact. You will be replaced.

Your daily reminder that AI is gonna replace your job. In this video, Klarna CEO Sebastian Siemeatkowski explains how they use OpenAI to replace around 20% of employees with AI every year. And that's just the beginning. An hour ago I posted the interview with OpenAI's CFO from Yahoo, OpenAI's goal is also to replace humans. Face this fact. You will be replaced.

Chubby♨️

208,100 Aufrufe • vor 1 Jahr

AI companies need humans to make their models better. They need people to create new ideas, conduct research, document discoveries and produce the knowledge that keeps AI useful. Cloudflare CEO Matthew Prince recently explained that AI companies are not looking for endless copies of information that already exists. They need humans to keep filling the gaps in knowledge, and he believes the people creating that value should be compensated for it. Because AI does not become intelligent in isolation. It learns from the work, knowledge and experiences people continuously put into the world. But under the current model, those contributions can make AI more valuable while the ownership and economic upside remain concentrated among a small number of companies and billionaires. The people helping educate AI risk becoming invisible in the process. Action Model is built around a different idea: The people contributing to AI should participate in its success. Through our community-owned Large Action Model, people can contribute real actions, workflows and knowledge, then earn ownership based on the value they help create. AI will continue learning from humanity. The value it creates should flow back to all the people making that intelligence possible, not just Big Tech companies.

AI companies need humans to make their models better. They need people to create new ideas, conduct research, document discoveries and produce the knowledge that keeps AI useful. Cloudflare CEO Matthew Prince recently explained that AI companies are not looking for endless copies of information that already exists. They need humans to keep filling the gaps in knowledge, and he believes the people creating that value should be compensated for it. Because AI does not become intelligent in isolation. It learns from the work, knowledge and experiences people continuously put into the world. But under the current model, those contributions can make AI more valuable while the ownership and economic upside remain concentrated among a small number of companies and billionaires. The people helping educate AI risk becoming invisible in the process. Action Model is built around a different idea: The people contributing to AI should participate in its success. Through our community-owned Large Action Model, people can contribute real actions, workflows and knowledge, then earn ownership based on the value they help create. AI will continue learning from humanity. The value it creates should flow back to all the people making that intelligence possible, not just Big Tech companies.

Action Model

10,723 Aufrufe • vor 4 Tagen

“I’m trying, I’m not good at English, what am I supposed to do? I’m not good. But I’m trying. And Scarlett, when she speaks, she’s so cute, but she has a really strong accent. She’s actually really sweet because she makes an effort to be understood. She speaks slowly and clearly when she talks to me. But I’m trying to learn English. I know I’m not doing a great job. I just wish I could learn it better. Still, we understand each other. That’s the important thing, we understand each other really well. Just… not through the language.” 🤧🤧

“I’m trying, I’m not good at English, what am I supposed to do? I’m not good. But I’m trying. And Scarlett, when she speaks, she’s so cute, but she has a really strong accent. She’s actually really sweet because she makes an effort to be understood. She speaks slowly and clearly when she talks to me. But I’m trying to learn English. I know I’m not doing a great job. I just wish I could learn it better. Still, we understand each other. That’s the important thing, we understand each other really well. Just… not through the language.” 🤧🤧

Korslayage

65,040 Aufrufe • vor 6 Monaten

Figma CEO Dylan Field Dylan Field says IQmaxxing is overrated: "All my experience with AI so far have shown that on formal domains, things that are verifiable, you can get to amazing outcomes with enough IQ points. And yet judgment is somehow not purely correlated with just adding IQ points to the model." "Design certainly does not seem to be correlated with adding IQ points to the model." "Humans being in the loop will be the way that you steer towards the best outcomes."

Figma CEO Dylan Field Dylan Field says IQmaxxing is overrated: "All my experience with AI so far have shown that on formal domains, things that are verifiable, you can get to amazing outcomes with enough IQ points. And yet judgment is somehow not purely correlated with just adding IQ points to the model." "Design certainly does not seem to be correlated with adding IQ points to the model." "Humans being in the loop will be the way that you steer towards the best outcomes."

sourcery

281,916 Aufrufe • vor 2 Monaten

🚨 BREAKING 🚨 Elon Musk just set a brutal deadline: “AI will surpass the combined intelligence of all humans in about 5 years.” After that, there may be almost nothing AI can’t do better than us. Except one thing: being human. We’re heading toward an era of unimaginable abundance — where everyone could have access to almost anything they can dream of. But there’s a darker side. If the intelligence gap between AI and humans becomes greater than the gap between humans and chimpanzees… Who’s really going to be in control? We were climbing trees and eating bananas not that long ago. Let’s see where we end up by 2036🤖

🚨 BREAKING 🚨 Elon Musk just set a brutal deadline: “AI will surpass the combined intelligence of all humans in about 5 years.” After that, there may be almost nothing AI can’t do better than us. Except one thing: being human. We’re heading toward an era of unimaginable abundance — where everyone could have access to almost anything they can dream of. But there’s a darker side. If the intelligence gap between AI and humans becomes greater than the gap between humans and chimpanzees… Who’s really going to be in control? We were climbing trees and eating bananas not that long ago. Let’s see where we end up by 2036🤖

DN_DEGEN

71,701 Aufrufe • vor 1 Monat