正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

Sam Altman says AI models will get exponentially more capable and cheaper per task History proves that as costs drop, demand explodes Even as models become efficient enough to run on our phones, the world's hunger for tokens will only increase "capitalism is doing its thing"

Haider.

64,661 subscribers

38,159 次观看 • 4 个月前 •via X (Twitter)

新闻政治科学技术

Anya Rossi• Live Now

Private livecam show

0 条评论

暂无评论

原始帖子的评论将显示在这里

相关视频

Avi Schiffmann, creator of the Friend necklace, says that the device is "like God" because of its omnipresence and superintelligence and this will only become more evident as AI models become sentient

Avi Schiffmann, creator of the Friend necklace, says that the device is "like God" because of its omnipresence and superintelligence and this will only become more evident as AI models become sentient

Tsarathustra

783,650 次观看 • 1 年前

Dylan Patel: “CPUs are sold out.” $INTC earnings and conference call confirmed the looming CPU shortage. There are three main drivers: - Agentic systems, as CPUs do the planning and orchestration. - Increasing complexity of and demand for reinforcement learning environments that run on CPUs. - Increasing deployment of AI instilled apps. Models run on GPUs but apps themselves run on CPUs. Demand for all these will grow secularly as AI systems will only get more agentic, RL environments will only get more complex, and deployment will only accelerate. As a result, GPU-to-CPU ratio in AI workloads will move from 1:8 to 1:2 or maybe even 1:1 in the future. CPU supercycle is about to start. $AMD $INTC $ARM

Dylan Patel: “CPUs are sold out.” $INTC earnings and conference call confirmed the looming CPU shortage. There are three main drivers: - Agentic systems, as CPUs do the planning and orchestration. - Increasing complexity of and demand for reinforcement learning environments that run on CPUs. - Increasing deployment of AI instilled apps. Models run on GPUs but apps themselves run on CPUs. Demand for all these will grow secularly as AI systems will only get more agentic, RL environments will only get more complex, and deployment will only accelerate. As a result, GPU-to-CPU ratio in AI workloads will move from 1:8 to 1:2 or maybe even 1:1 in the future. CPU supercycle is about to start. $AMD $INTC $ARM

Oguz Erkan

93,419 次观看 • 1 个月前

Sam Altman says each increase in capability and decrease in cost creates massive new demand "even 30 GW with today's models would saturate fast" If GPT-6 is 30 IQ points above GPT-5 and can work for days to months, the economic value will skyrocket

Sam Altman says each increase in capability and decrease in cost creates massive new demand "even 30 GW with today's models would saturate fast" If GPT-6 is 30 IQ points above GPT-5 and can work for days to months, the economic value will skyrocket

Haider.

213,352 次观看 • 8 个月前

Sam Altman says an AI glut is inevitable, maybe in 2-3 years, maybe 5-6, and likely more than once Costs have fallen 40x per year per unit of intelligence, which strains infrastructure "if breakthroughs continue, personal AGIs could run on laptops" We expect demand to grow, but some people will get burned

Sam Altman says an AI glut is inevitable, maybe in 2-3 years, maybe 5-6, and likely more than once Costs have fallen 40x per year per unit of intelligence, which strains infrastructure "if breakthroughs continue, personal AGIs could run on laptops" We expect demand to grow, but some people will get burned

Haider.

977,618 次观看 • 7 个月前

Dario Amodei says AI products start to saturate when the models become too capable Chatbots will remain useful, but the real gains from smarter models are now showing up more clearly in coding tools and agentic systems "in this industry, there's no such thing as a stable product"

Dario Amodei says AI products start to saturate when the models become too capable Chatbots will remain useful, but the real gains from smarter models are now showing up more clearly in coding tools and agentic systems "in this industry, there's no such thing as a stable product"

Haider.

33,497 次观看 • 1 个月前

Sam Altman says the leap from GPT-4 to GPT-5 will be as big as that of GPT-3 to 4 and the plan is to integrate the GPT and o series of models into one model that can do everything

Sam Altman says the leap from GPT-4 to GPT-5 will be as big as that of GPT-3 to 4 and the plan is to integrate the GPT and o series of models into one model that can do everything

Tsarathustra

224,102 次观看 • 1 年前

Sam Altman says even as AI becomes unbelievably intelligent, it will not become the center of the human story "we're wired to care about people, not machines" And we will continue to value human achievements and connections far more than machine performance

Sam Altman says even as AI becomes unbelievably intelligent, it will not become the center of the human story "we're wired to care about people, not machines" And we will continue to value human achievements and connections far more than machine performance

Haider.

228,747 次观看 • 8 个月前

Sam Altman: "I bet there is another new architecture to find" Sam Altman believes we are on the verge of discovering a new underlying architecture that will be as big of a leap forward as Transformers were over LSTMs. He noted that we finally have AI models that are smart enough to help conduct this level of research (GPT 5.4 and above 👀) His direct advice to builders looking for the next major leap is to look for a "mega breakthrough" and use current models to help them find it.

Sam Altman: "I bet there is another new architecture to find" Sam Altman believes we are on the verge of discovering a new underlying architecture that will be as big of a leap forward as Transformers were over LSTMs. He noted that we finally have AI models that are smart enough to help conduct this level of research (GPT 5.4 and above 👀) His direct advice to builders looking for the next major leap is to look for a "mega breakthrough" and use current models to help them find it.

Chris

289,423 次观看 • 3 个月前

"I think one of the truths is that the demand for AI intelligence and smarter AI intelligence is going to be insatiable. Some people disagree that once we reach certain thresholds then you don't need more intelligence. To that I ask people, 'Have you ever worked with or managed someone in a work environment and wouldn't press a button to make them smarter?'" George elaborates on his thesis that the demand for smarter AI intelligence will only increase over time even as models get more powerful. George Cameron

"I think one of the truths is that the demand for AI intelligence and smarter AI intelligence is going to be insatiable. Some people disagree that once we reach certain thresholds then you don't need more intelligence. To that I ask people, 'Have you ever worked with or managed someone in a work environment and wouldn't press a button to make them smarter?'" George elaborates on his thesis that the demand for smarter AI intelligence will only increase over time even as models get more powerful. George Cameron

Latent.Space

10,629 次观看 • 5 个月前

$META is betting on $AMD for AI inference. As AI models become more commoditized, total cost of ownership for token generation matters more than ever. $AMD GPUs are far more cost-efficient than $NVDA for inference. With 3.5B daily users ready to use its AI models, $META is going all-in on $AMD

$META is betting on $AMD for AI inference. As AI models become more commoditized, total cost of ownership for token generation matters more than ever. $AMD GPUs are far more cost-efficient than $NVDA for inference. With 3.5B daily users ready to use its AI models, $META is going all-in on $AMD

Daniel

20,558 次观看 • 4 个月前

We are in for a wild ride and we barely started. Sam Altman: "In 2 years I think the models will be doing bigger chunks of science and making important discoveries and that is a crazy thing that will have a significant impact on the world"

We are in for a wild ride and we barely started. Sam Altman: "In 2 years I think the models will be doing bigger chunks of science and making important discoveries and that is a crazy thing that will have a significant impact on the world"

Chubby♨️

13,371 次观看 • 8 个月前

Sam Altman says the leap from GPT-4 to GPT-5 will be as big as that of GPT-3 to 4 and the plan is to integrate the GPT and o series of models into one model that can do everything - that’s the AGI

Sam Altman says the leap from GPT-4 to GPT-5 will be as big as that of GPT-3 to 4 and the plan is to integrate the GPT and o series of models into one model that can do everything - that’s the AGI

Chubby♨️

231,778 次观看 • 1 年前

Today we’re announcing a finding that breaks a core assumption in AI: that bigger models are harder to understand. We show the opposite. When interpretability is built into training, models become MORE understandable as they become more capable.

Today we’re announcing a finding that breaks a core assumption in AI: that bigger models are harder to understand. We show the opposite. When interpretability is built into training, models become MORE understandable as they become more capable.

Guide Labs

13,329 次观看 • 11 天前

Uber CEO Dara Khosrowshahi: "We Blew Through Our AI Budget in a Quarter" In response, Uber is doing three things: -Slowing headcount growth as AI makes engineers more productive -Reserving expensive frontier models for exploration only and shifting production workloads to cheaper open source models -Capping each employee at $1,500/month per AI coding tool

Uber CEO Dara Khosrowshahi: "We Blew Through Our AI Budget in a Quarter" In response, Uber is doing three things: -Slowing headcount growth as AI makes engineers more productive -Reserving expensive frontier models for exploration only and shifting production workloads to cheaper open source models -Capping each employee at $1,500/month per AI coding tool

Wall Street Rollup

99,031 次观看 • 19 天前

Sam Altman says Codex isn't far from week-long tasks and is already doing day-long runs The jump in task length feels disorientingly fast Reaching week-long tasks will require smarter models, longer context, and better memory

Sam Altman says Codex isn't far from week-long tasks and is already doing day-long runs The jump in task length feels disorientingly fast Reaching week-long tasks will require smarter models, longer context, and better memory

Haider.

69,294 次观看 • 8 个月前

Mustafa Suleyman says as more compute is added to AI models they become easier to control, steer and align

Mustafa Suleyman says as more compute is added to AI models they become easier to control, steer and align

Tsarathustra

25,389 次观看 • 1 年前

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?

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

235,118 次观看 • 1 年前

Sam Altman: Imagine models 100x more capable, with 100x context, speed, & coherence. Tell us what to build! We're crafting the perfect AI tools for you. "We are gonna get there"

Sam Altman: Imagine models 100x more capable, with 100x context, speed, & coherence. Tell us what to build! We're crafting the perfect AI tools for you. "We are gonna get there"

Chubby♨️

52,313 次观看 • 4 个月前