Video wird geladen...

Video konnte nicht geladen werden

Zur Startseite

"AI risk is moving into training itself." Sam Altman, says capability gains are exposing new alignment risks before models are even deployed. AI safety is becoming a training problem, not just a deployment problem.

JonhernandezIA's profile picture

Jon Hernandez

26,080 subscribers

13,430 Aufrufe • vor 22 Tagen •via X (Twitter)

LIVE

Anya Rossi• Live Now

Private livecam show

10 Kommentare

Profilbild von Alex Heath
Alex Heathvor 22 Tagen

Hey - super not cool to crop in like this to hide my logos

Profilbild von Jon Hernandez
Jon Hernandezvor 22 Tagen

Source:

Profilbild von Rupert Davies
Rupert Daviesvor 22 Tagen

By the very nature of its architecture, if risk moves into training, the objectives were never right. You don't bolt alignment on after deployment.

Profilbild von Maximiliano Torres
Maximiliano Torresvor 22 Tagen

En pocas palabras crean el asteroide que va achocar con la tierra y nos piden que ayudemos entre todos a construir las naves para ir a Marte y salvarnos

Profilbild von Jannjhay de Leon
Jannjhay de Leonvor 22 Tagen

They won't ^_^ they will be mirrors

Profilbild von Bete Moreira
Bete Moreiravor 22 Tagen

Fico observando um ser humano DESSES falando de uma máquina e penso que ele está completamente fora de orbita. Sinceramente é lamentável vê em que um ser humano DESSES devota sua vida, lembrando que ele cavou um buraco no chão pensando em se esconder. Se esconder de quem?

Profilbild von Ant A. 🇺🇸
Ant A. 🇺🇸vor 22 Tagen

@Grok ASI doing everybody’s job where no one is screwed will be the only way.

Profilbild von RIGHT ON "Words of Wisdom"
RIGHT ON "Words of Wisdom"vor 22 Tagen

AI REGULATIONS ARE CRUCIAL! JUST ASK THE ALIENS.

Profilbild von manzi
manzivor 21 Tagen

Sam Altman. which habit still needs a person?

Profilbild von PoopooDemon
PoopooDemonvor 22 Tagen

Alignment arguably will matter as much if not more than raw capabilities with AGI and above AI models.

Ähnliche Videos

📁 Sam Altman, CEO of OpenAI, says the problem with AI is not if it works. It is what happens when it does. Because curing disease is not enough. The real question is: what do you do when you are no longer needed?
2:07

📁 Sam Altman, CEO of OpenAI, says the problem with AI is not if it works. It is what happens when it does. Because curing disease is not enough. The real question is: what do you do when you are no longer needed?

Jon Hernandez

49,677 Aufrufe • vor 5 Monaten

Sam Altman says AI may need a new hardware form factor OpenAI and Jony Ive are exploring a family of devices to make AI easier to use — this will take time (not this year, maybe not next) The Goal is a new kind of computer built around AI, a companion through your life
0:39

Sam Altman says AI may need a new hardware form factor OpenAI and Jony Ive are exploring a family of devices to make AI easier to use — this will take time (not this year, maybe not next) The Goal is a new kind of computer built around AI, a companion through your life

Haider.

193,273 Aufrufe • vor 11 Monaten

Sam Altman says AI models are general-purpose reasoning engines with a huge world knowledge built into them The next milestone is letting a model see something totally new once, then use it reliably and learn it faster than humans "that doesn't feel very far away"
0:59

Sam Altman says AI models are general-purpose reasoning engines with a huge world knowledge built into them The next milestone is letting a model see something totally new once, then use it reliably and learn it faster than humans "that doesn't feel very far away"

Haider.

20,500 Aufrufe • vor 7 Monaten

Richard Sutton says AI safety advocates are creating the opposite of what they seek - we should not try to solve the AI control problem because a decentralized system of independent agents is better than everything being aligned under one goal
0:48

Richard Sutton says AI safety advocates are creating the opposite of what they seek - we should not try to solve the AI control problem because a decentralized system of independent agents is better than everything being aligned under one goal

Tsarathustra

136,399 Aufrufe • vor 2 Jahren

Aravind Srinivas just explained why China’s open-source AI may become more powerful than ever. And why Anthropic has lobbied very hard for export control. "the only reason why there is even a 12-month gap between open source and frontier models is export controls. But there is a chance that, because of that, they now get really good at the physical layer. One advantage they (China) have is that they can actually build data centers a lot faster. Power is not a problem. Permits are not a problem. People are not a problem. Labor is not a problem. Expertise is not a problem. And so, by forcing them to go out there and build all this, you are converting them into a far more potent competitor." --- From "20VC with Harry Stebbings" YouTube channel (Harry Stebbings ), link in comment
0:57

Aravind Srinivas just explained why China’s open-source AI may become more powerful than ever. And why Anthropic has lobbied very hard for export control. "the only reason why there is even a 12-month gap between open source and frontier models is export controls. But there is a chance that, because of that, they now get really good at the physical layer. One advantage they (China) have is that they can actually build data centers a lot faster. Power is not a problem. Permits are not a problem. People are not a problem. Labor is not a problem. Expertise is not a problem. And so, by forcing them to go out there and build all this, you are converting them into a far more potent competitor." --- From "20VC with Harry Stebbings" YouTube channel (Harry Stebbings ), link in comment

Rohan Paul

139,116 Aufrufe • vor 2 Monaten

"The real problem at the moment is that the technology is intrinsically unsafe." Stuart Russell, UC Berkeley AI professor and president of IASEAI, says we are scaling AI systems we still don't fully understand. Making AI more powerful is not the same as making it controllable. Safety requires understanding the systems we are building.
2:30

"The real problem at the moment is that the technology is intrinsically unsafe." Stuart Russell, UC Berkeley AI professor and president of IASEAI, says we are scaling AI systems we still don't fully understand. Making AI more powerful is not the same as making it controllable. Safety requires understanding the systems we are building.

Jon Hernandez

10,415 Aufrufe • vor 3 Tagen

The gap between what AI models can do and what products actually exist is massive right now.⁣ ⁣ Sam Altman says reasoning models opened up entirely new possibilities that barely anyone is exploring yet.⁣ ⁣ Tomorrow's products are yet to be built. Will you build them?
1:31

The gap between what AI models can do and what products actually exist is massive right now.⁣ ⁣ Sam Altman says reasoning models opened up entirely new possibilities that barely anyone is exploring yet.⁣ ⁣ Tomorrow's products are yet to be built. Will you build them?

Garry Tan

235,164 Aufrufe • vor 1 Jahr

AI is shaping the future whether we like it or not. Technology itself is not the problem. The problem is deploying powerful tools without guardrails, transparency, or accountability. AI should be a tool that works for people – not one that extracts value and burdens workers.
2:03

AI is shaping the future whether we like it or not. Technology itself is not the problem. The problem is deploying powerful tools without guardrails, transparency, or accountability. AI should be a tool that works for people – not one that extracts value and burdens workers.

Rep. Ilhan Omar

28,248 Aufrufe • vor 7 Monaten

This Sam Altman story is chilling. He says the current AI moment feels like the final night before the world realized COVID would change everything. A few people saw what was coming. Everyone else kept partying. That is exactly what AI feels like now. We are not waiting for the disruption. We are living in the last moments before everyone notices it.
1:57

This Sam Altman story is chilling. He says the current AI moment feels like the final night before the world realized COVID would change everything. A few people saw what was coming. Everyone else kept partying. That is exactly what AI feels like now. We are not waiting for the disruption. We are living in the last moments before everyone notices it.

VraserX e/acc

82,217 Aufrufe • vor 5 Monaten

Anthropic CEO Dario Amodei says AI companies need to work together on safety standards, including potentially the pace at which new models are released with the government “in the room.” CBS News’ Jo Ling Kent says Amodei is willing to work with competitors including OpenAI’s Sam Altman, who has backed Amodei’s push for independent safety testing.
3:31

Anthropic CEO Dario Amodei says AI companies need to work together on safety standards, including potentially the pace at which new models are released with the government “in the room.” CBS News’ Jo Ling Kent says Amodei is willing to work with competitors including OpenAI’s Sam Altman, who has backed Amodei’s push for independent safety testing.

Face The Nation

12,125 Aufrufe • vor 11 Tagen

"the only reason why there is even a 12-month gap between open source and frontier models is US export controls." Perplexity’s CEO Aravind Srinivas on why China’s open-source AI may become more powerful than ever. And why Anthropic has lobbied very hard for export control. "There is a chance that, because of the export controls, they now get really good at the physical layer. One advantage they (China) have is that they can actually build data centers a lot faster. Power is not a problem. Permits are not a problem. People are not a problem. Labor is not a problem. Expertise is not a problem. And so, by forcing them to go out there and build all this, you are converting them into a far more potent competitor." --- From "20VC with Harry Stebbings" YouTube channel (Harry Stebbings ), link in comment
0:57

"the only reason why there is even a 12-month gap between open source and frontier models is US export controls." Perplexity’s CEO Aravind Srinivas on why China’s open-source AI may become more powerful than ever. And why Anthropic has lobbied very hard for export control. "There is a chance that, because of the export controls, they now get really good at the physical layer. One advantage they (China) have is that they can actually build data centers a lot faster. Power is not a problem. Permits are not a problem. People are not a problem. Labor is not a problem. Expertise is not a problem. And so, by forcing them to go out there and build all this, you are converting them into a far more potent competitor." --- From "20VC with Harry Stebbings" YouTube channel (Harry Stebbings ), link in comment

Rohan Paul

34,754 Aufrufe • vor 4 Tagen

Sam Altman says true superintelligence already exists, but it's not a single brain It is the collective "scaffolding" of human civilization If AI discovers new physics that humans later use, is that "AI as superintelligence" or just the scaffold getting taller? I don't know
0:57

Sam Altman says true superintelligence already exists, but it's not a single brain It is the collective "scaffolding" of human civilization If AI discovers new physics that humans later use, is that "AI as superintelligence" or just the scaffold getting taller? I don't know

Haider.

75,051 Aufrufe • vor 10 Monaten

A new startup in Silicon Valley is already valued at $10 billion, it’s using artificial intelligence to replace workers They’re training AI to replace doctors, lawyers, bankers and even jobs like recommending wine Entire industries will be replaced with ai workers “AI labs are hiring large armies of people to help create these data sets and teaching models how to learn from them” “Are these people not kind of training their future replacements?” (Yes, of course they are)
2:52

A new startup in Silicon Valley is already valued at $10 billion, it’s using artificial intelligence to replace workers They’re training AI to replace doctors, lawyers, bankers and even jobs like recommending wine Entire industries will be replaced with ai workers “AI labs are hiring large armies of people to help create these data sets and teaching models how to learn from them” “Are these people not kind of training their future replacements?” (Yes, of course they are)

Wall Street Apes

243,266 Aufrufe • vor 7 Monaten

Sam Altman says abundant intelligence and abundant energy will lead to solving every problem in physics and eventually we'll be talking not about using nuclear fusion or solar to power AI but a Dyson Sphere
1:30

Sam Altman says abundant intelligence and abundant energy will lead to solving every problem in physics and eventually we'll be talking not about using nuclear fusion or solar to power AI but a Dyson Sphere

Tsarathustra

86,892 Aufrufe • vor 1 Jahr