Video wird geladen...

Video konnte nicht geladen werden

Zur Startseite

Dario Amodei says pre-training sits somewhere between learning and evolution. Humans inherit priors shaped over millions of years. LLMs start as random weights and distill trillions of tokens into those priors. We describe them using human learning metaphors. But the analogy only goes so far.

vitrupo's profile picture

vitrupo

35,430 subscribers

45,540 Aufrufe • vor 4 Monaten •via X (Twitter)

BildungWissenschaft & Technologie
LIVE

Anya Rossi• Live Now

Private livecam show

0 Kommentare

Keine Kommentare verfügbar

Kommentare vom Original-Post werden hier angezeigt

Ähnliche Videos

David Silver says human data is AI's fossil fuel. "We mine it and burn it in our LLMs." It gave us a head start -- but it will run out. The sustainable path is reinforcement learning: systems that learn from experience, generate more, and keep learning.
0:56

David Silver says human data is AI's fossil fuel. "We mine it and burn it in our LLMs." It gave us a head start -- but it will run out. The sustainable path is reinforcement learning: systems that learn from experience, generate more, and keep learning.

vitrupo

20,190 Aufrufe • vor 1 Jahr

Anthropic CEO Dario Amodei says if exponential progress continues for 1–3 years, AI will cross the frontier of human knowledge. A very small number of humans with swarms of AI agents will drive new scientific discoveries and economic activity. “Then things really go crazy.”
1:14

Anthropic CEO Dario Amodei says if exponential progress continues for 1–3 years, AI will cross the frontier of human knowledge. A very small number of humans with swarms of AI agents will drive new scientific discoveries and economic activity. “Then things really go crazy.”

vitrupo

42,131 Aufrufe • vor 10 Monaten

DreamPhysics: Learning Physical Properties of Dynamic 3D Gaussians from Video Diffusion Priors Code:
0:21

DreamPhysics: Learning Physical Properties of Dynamic 3D Gaussians from Video Diffusion Priors Code:

MrNeRF

12,360 Aufrufe • vor 2 Jahren

AI training faces a very different set of trade-offs compared to human evolution. Because you can directly copy a trained model basically for free, it makes sense to amortize in the pre-trained weights the learning that humans spread across a lifetime.
2:59

AI training faces a very different set of trade-offs compared to human evolution. Because you can directly copy a trained model basically for free, it makes sense to amortize in the pre-trained weights the learning that humans spread across a lifetime.

Dwarkesh Patel

25,747 Aufrufe • vor 2 Monaten

Andrej Karpathy says when we sleep, our daily “context window” is distilled into the weights of the mind. AI doesn't dream like us yet. But as we teach machines to reflect and compress experience, we may find both humans and AI converging on the same cognitive architecture.
1:20

Andrej Karpathy says when we sleep, our daily “context window” is distilled into the weights of the mind. AI doesn't dream like us yet. But as we teach machines to reflect and compress experience, we may find both humans and AI converging on the same cognitive architecture.

vitrupo

592,769 Aufrufe • vor 8 Monaten

Dario Amodei says future AI models will make far fewer mistakes than humans, but stranger mistakes Humans make predictable errors; AI’s mistakes are subtle, coherent, and deceptive. "you can tell when a human is wrong -- but AI mistakes sound just as confident as the truth"
1:01

Dario Amodei says future AI models will make far fewer mistakes than humans, but stranger mistakes Humans make predictable errors; AI’s mistakes are subtle, coherent, and deceptive. "you can tell when a human is wrong -- but AI mistakes sound just as confident as the truth"

Haider.

41,431 Aufrufe • vor 11 Monaten

Dario Amodei says the spark was seeing engineers stop writing code and start supervising it. Inside Anthropic, Claude writes much of the software, and humans review like editors. The twist is meta: some of that code helps build the next Claude. That’s the self-improvement loop tightening, says Dario Amodei.
0:49

Dario Amodei says the spark was seeing engineers stop writing code and start supervising it. Inside Anthropic, Claude writes much of the software, and humans review like editors. The twist is meta: some of that code helps build the next Claude. That’s the self-improvement loop tightening, says Dario Amodei.

Wes Roth

87,585 Aufrufe • vor 5 Monaten

Jim Fan says NVIDIA trained humanoid robots to walk and move like humans -- zero-shot transfer from simulation to the real world. 10 years of learning, compressed into 2 hours of simulation. Turns out you don’t need a giant model to master motion. “1.5 million parameters, not billion, to capture the subconscious processing of the human body.” Big embodiment, small brain.
1:26

Jim Fan says NVIDIA trained humanoid robots to walk and move like humans -- zero-shot transfer from simulation to the real world. 10 years of learning, compressed into 2 hours of simulation. Turns out you don’t need a giant model to master motion. “1.5 million parameters, not billion, to capture the subconscious processing of the human body.” Big embodiment, small brain.

vitrupo

714,670 Aufrufe • vor 1 Jahr

François Chollet says most definitions of AGI are really definitions of automation. AGI isn’t about doing tasks. It’s about learning new ones as efficiently as a human. We may automate the economy before we build intelligence.
0:59

François Chollet says most definitions of AGI are really definitions of automation. AGI isn’t about doing tasks. It’s about learning new ones as efficiently as a human. We may automate the economy before we build intelligence.

vitrupo

11,327 Aufrufe • vor 3 Monaten

Geoffrey Hinton says LLMs are immortal. Destroy the hardware, save the weights. Rebuild them later, and “the very same being comes back to life.” But humans don't work that way. We're bound to our substrate. Not even Kurzweil gets a backup.
1:28

Geoffrey Hinton says LLMs are immortal. Destroy the hardware, save the weights. Rebuild them later, and “the very same being comes back to life.” But humans don't work that way. We're bound to our substrate. Not even Kurzweil gets a backup.

vitrupo

323,189 Aufrufe • vor 11 Monaten

DeepMind's Carolina Parada says Asimov's Three Laws aren't enough to keep robots safe. Her team built the Asimov Dataset, a growing set of safety priors, extracted by AI and refined by humans. Robots don't follow rules. They learn to reason about harm.
1:51

DeepMind's Carolina Parada says Asimov's Three Laws aren't enough to keep robots safe. Her team built the Asimov Dataset, a growing set of safety priors, extracted by AI and refined by humans. Robots don't follow rules. They learn to reason about harm.

vitrupo

73,264 Aufrufe • vor 1 Jahr

Learning Temporally Consistent Video Depth from Video Diffusion Priors This work addresses the challenge of video depth estimation, which expects not only per-frame accuracy but, more importantly, cross-frame consistency. Instead of directly
1:25

Learning Temporally Consistent Video Depth from Video Diffusion Priors This work addresses the challenge of video depth estimation, which expects not only per-frame accuracy but, more importantly, cross-frame consistency. Instead of directly

AK

81,634 Aufrufe • vor 2 Jahren

this was the most important bit of the ilya interview imo. humans are not AGI, and LLMs won’t become like humans with more scaling. humans lack a huge amount of knowledge and have insanely good “continual learning” powered by extremely good reward functions (except for addiction) that guide them. and so it’s more about the learning rather than knowledge. the term AGI was coined as a response to the narrow AI which was only capable of intelligence in a specific domain (chess, game, code, etc) but General intelligence is able to learn different topics because that’s what “learning” is about. pre-training was thought to be the solution to reaching AGI because you train AI on so many different topics. but they have “overshot” that target. amazing perspective.
1:57

this was the most important bit of the ilya interview imo. humans are not AGI, and LLMs won’t become like humans with more scaling. humans lack a huge amount of knowledge and have insanely good “continual learning” powered by extremely good reward functions (except for addiction) that guide them. and so it’s more about the learning rather than knowledge. the term AGI was coined as a response to the narrow AI which was only capable of intelligence in a specific domain (chess, game, code, etc) but General intelligence is able to learn different topics because that’s what “learning” is about. pre-training was thought to be the solution to reaching AGI because you train AI on so many different topics. but they have “overshot” that target. amazing perspective.

ℏεsam

53,442 Aufrufe • vor 7 Monaten

The evolution of the relationship between humans and electronic devices 1910-2030 using Al.
0:30

The evolution of the relationship between humans and electronic devices 1910-2030 using Al.

Historic Vids

4,588,023 Aufrufe • vor 3 Jahren

The evolution of the relationship between humans and electronic devices 1910 - 2030 using AI
0:31

The evolution of the relationship between humans and electronic devices 1910 - 2030 using AI

Historic Vids

812,440 Aufrufe • vor 2 Jahren

Very excited to announce: Keypoint Action Tokens! We found that LLMs can be repurposed as "imitation learning engines" for robots, by representing both observations & actions as 3D keypoints, and feeding into an LLM for in-context learning. See: More 👇
0:29

Very excited to announce: Keypoint Action Tokens! We found that LLMs can be repurposed as "imitation learning engines" for robots, by representing both observations & actions as 3D keypoints, and feeding into an LLM for in-context learning. See: More 👇

Edward Johns

32,569 Aufrufe • vor 2 Jahren

Over the years, DU has emerged as a prime centre of learning and innovation.
3:33

Over the years, DU has emerged as a prime centre of learning and innovation.

Narendra Modi

437,824 Aufrufe • vor 3 Jahren

Eric Schmidt says war is no longer constrained by humans. It's AI vs AI. “No human can plan a battle… without reinforcement learning.” The real conflict lives in data centers and drone swarms. No human can follow it in real time.
1:17

Eric Schmidt says war is no longer constrained by humans. It's AI vs AI. “No human can plan a battle… without reinforcement learning.” The real conflict lives in data centers and drone swarms. No human can follow it in real time.

vitrupo

112,802 Aufrufe • vor 1 Jahr