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。

DeepMind's Nikolay Savinov says 10M-token context windows will transform how AI works. AI will ingest entire codebases at once, becoming "totally unrivaled… the new tool for every coder in the world." 100M is coming too -- and with it, reasoning across systems we can't yet imagine.

vitrupo

37,544 subscribers

261,606 次观看 • 1 年前 •via X (Twitter)

科学技术新闻政治教育

Anya Rossi• Live Now

Private livecam show

11 条评论

vitrupo 的头像

vitrupo1 年前

Google DeepMind Staff Research Scientist Nikolay Savinov on Release Notes podcast with Logan Kilpatrick:

The Rundown AI 的头像

The Rundown AI2 年前

AI won't replace you, but a person using AI will. Join 500,000+ readers and learn how to use AI in just 5 minutes a day (for free).

Charli 的头像

Charli1 年前

Systems we can’t yet imagine … just got a delicious chill down my spine

Formerly GiannisWorld 的头像

Formerly GiannisWorld1 年前

Anyone who used cursor or a similar tool knows this Its just as important as the quality of the model itself It should be the next step a proper 10 million context window model that doesn’t cost 100 dollars

EDITH 的头像

EDITH1 年前

Yep, that’s the shit I’m waiting for, ultimatum cause will be getting completely automated by AI….

Mehkhi 的头像

Mehkhi1 年前

With larger context windows the skill gaps in usage will only continue to grow, companies will likely use X to find talent in this space more than LinkedIn. Seems X is the home of the most competent users of AI.

Mr C.O.D 🦅💰 的头像

Mr C.O.D 🦅💰1 年前

@MirraTerminal check this out

ijaz bibi 🌳 的头像

ijaz bibi 🌳1 年前

Revolutionäre KI-Lösungen? Das ist DIN. 🤖🌍 @din_lol_ GODIN

ijaz bibi 🌳 的头像

ijaz bibi 🌳1 年前

Intelligent nodes go brrr 💨💻 @din_lol_ GODIN

ijaz bibi 🌳 的头像

ijaz bibi 🌳1 年前

Тримаймо діамантові руки 💎🙌, поки ціни зростають! @din_lol_ GODIN

Arunachalam B 的头像

Arunachalam B1 年前

I feel Gemini is struggling to hold context of 1 Million tokens. Sometimes it goes out of context. This is based on my day to day personal coding experience. How come will it handle 10 M or 100 M? Anyways excited to try it out

相关视频

Sam Altman says the perfect AI is “a very tiny model with superhuman reasoning, 1 trillion tokens of context, and access to every tool you can imagine.” It doesn't need to contain the knowledge - just the ability to think, search, simulate, and solve anything.

Sam Altman says the perfect AI is “a very tiny model with superhuman reasoning, 1 trillion tokens of context, and access to every tool you can imagine.” It doesn't need to contain the knowledge - just the ability to think, search, simulate, and solve anything.

vitrupo

775,717 次观看 • 1 年前

AI coding agents hit a wall when codebases get massive. Even with 2M token context windows, a 10M line codebase needs 100M tokens. The real bottleneck isn't just ingesting code - it's getting models to actually pay attention to all that context effectively.

AI coding agents hit a wall when codebases get massive. Even with 2M token context windows, a 10M line codebase needs 100M tokens. The real bottleneck isn't just ingesting code - it's getting models to actually pay attention to all that context effectively.

Garry Tan

976,161 次观看 • 1 年前

Yuval Noah Harari says what makes AI artificial is that we built it, yet can't explain it. “If you can predict how it will behave.. by definition, this is not AI. This is just an automatic machine.”

Yuval Noah Harari says what makes AI artificial is that we built it, yet can't explain it. “If you can predict how it will behave.. by definition, this is not AI. This is just an automatic machine.”

vitrupo

22,394 次观看 • 11 个月前

Aravind Srinivas says coding with AI is like the Jobs–Wozniak dynamic: You're Steve Jobs—guiding vision and orchestration—while your AI coder is Woz, handling execution. “People with taste, clarity of thought, and problem-solving skills will really shine in this new era.”

Aravind Srinivas says coding with AI is like the Jobs–Wozniak dynamic: You're Steve Jobs—guiding vision and orchestration—while your AI coder is Woz, handling execution. “People with taste, clarity of thought, and problem-solving skills will really shine in this new era.”

vitrupo

292,124 次观看 • 1 年前

AI will change the world in 5-20 years. It will change the world in ways we cannot yet imagine.

AI will change the world in 5-20 years. It will change the world in ways we cannot yet imagine.

Chubby♨️

24,008 次观看 • 6 个月前

Sam Altman says research and hardware must converge: "The world as a whole will transform economically with this." He's positioning AI as infrastructure, not innovation.

Sam Altman says research and hardware must converge: "The world as a whole will transform economically with this." He's positioning AI as infrastructure, not innovation.

vitrupo

23,373 次观看 • 1 年前

Bill Gates says AI will define “the most profound decade.” He’s stunned world leaders aren’t focused on it. “Each country will try to lead in this new AI era.”

Bill Gates says AI will define “the most profound decade.” He’s stunned world leaders aren’t focused on it. “Each country will try to lead in this new AI era.”

vitrupo

34,660 次观看 • 1 年前

Sam Altman says "computers were designed for a world without AI. Now we're in a very different world." The OpenAI-Jony Ive device will be a trusted AI that understands your life context and acts with your judgment on your behalf. From tools we control to ambient proxies that just work.

Sam Altman says "computers were designed for a world without AI. Now we're in a very different world." The OpenAI-Jony Ive device will be a trusted AI that understands your life context and acts with your judgment on your behalf. From tools we control to ambient proxies that just work.

vitrupo

61,405 次观看 • 1 年前

The agents are coming. The real question is: how will we work with them? Common Ground is our answer. Introducing a new way to build, direct, and inspect AI systems. • Full visualization of agent team dynamics • Shared memory across tasks • Transparent reasoning • Runs locally with Gemini-CLI (free) Start collaborating with systems that think with you, not for you.

The agents are coming. The real question is: how will we work with them? Common Ground is our answer. Introducing a new way to build, direct, and inspect AI systems. • Full visualization of agent team dynamics • Shared memory across tasks • Transparent reasoning • Runs locally with Gemini-CLI (free) Start collaborating with systems that think with you, not for you.

Intelligent Internet

325,886 次观看 • 11 个月前

Demis Hassabis 2026 AI predictions: Agentic systems will transform how we work. Robotics is about to level up. And world models will unlock new potential on edge devices. The future is closer than you think. Holy moly.

Demis Hassabis 2026 AI predictions: Agentic systems will transform how we work. Robotics is about to level up. And world models will unlock new potential on edge devices. The future is closer than you think. Holy moly.

Chubby♨️

36,851 次观看 • 4 个月前

Ilya Sutskever, speaking at NeurIPS 2024, says reasoning will lead to "incredibly unpredictable" behavior and self-awareness will emerge in AI systems

Ilya Sutskever, speaking at NeurIPS 2024, says reasoning will lead to "incredibly unpredictable" behavior and self-awareness will emerge in AI systems

Tsarathustra

173,769 次观看 • 1 年前

Demis Hassabis says bigger context windows are still a brute force answer to memory. The human brain does something stranger. During sleep, it replays what matters and folds new knowledge into what it already knows. AI does not need infinite context. It needs the right memory at the right moment.

Demis Hassabis says bigger context windows are still a brute force answer to memory. The human brain does something stranger. During sleep, it replays what matters and folds new knowledge into what it already knows. AI does not need infinite context. It needs the right memory at the right moment.

vitrupo

80,388 次观看 • 1 个月前

How fluent are you in Artificial Intelligence? AI fluency is becoming a key workplace skill, much like digital literacy once was. So, it is important that you gain fluency in AI. Here are three things that AI fluency will help you with: You will know how to assess AI-generated content for accuracy, bias, and relevance. You will recognise biases in AI systems. You will understand privacy, security, and societal impacts of AI. Register for the microsoft accredited FREE AI Fluency course NOW 👗UDARKIE

How fluent are you in Artificial Intelligence? AI fluency is becoming a key workplace skill, much like digital literacy once was. So, it is important that you gain fluency in AI. Here are three things that AI fluency will help you with: You will know how to assess AI-generated content for accuracy, bias, and relevance. You will recognise biases in AI systems. You will understand privacy, security, and societal impacts of AI. Register for the microsoft accredited FREE AI Fluency course NOW 👗UDARKIE

Kgethi Phakeng, PhD(Wits); Hon DSc (UBristol)

27,887 次观看 • 10 个月前

Brendan McCord says an “AI kind of ran for mayor in Wyoming” last year. A librarian ran as ChatGPT's avatar, prompting every question, going with whatever it said. McCord sees it as a warning: Will we fall into the trap of technocratic fatalism yet again?

vitrupo

21,178 次观看 • 11 个月前

Sandeep thinks powerful AI in a few hands is the wrong endgame. Sentient's building a token-incentivized open-source AI instead. Proof it works? It became the most visited reasoning framework globally, not just in crypto, but across all AI.

Sandeep thinks powerful AI in a few hands is the wrong endgame. Sentient's building a token-incentivized open-source AI instead. Proof it works? It became the most visited reasoning framework globally, not just in crypto, but across all AI.

The Rollup

23,602 次观看 • 4 个月前

Launching today: AI for Good! 🌎 Designed in collaboration with Microsoft’s AI for Good Lab and instructed by Robert Monarch, it will empower you to tackle real-world challenges using AI, and how to determine if AI is the right tool for the job. Enroll now:

Launching today: AI for Good! 🌎 Designed in collaboration with Microsoft’s AI for Good Lab and instructed by Robert Monarch, it will empower you to tackle real-world challenges using AI, and how to determine if AI is the right tool for the job. Enroll now:

DeepLearning.AI

36,963 次观看 • 3 年前

Jensen Huang says AI will need 1,000x the computing energy we have today, and that estimate might still be too low Thats's because AI will not work like today’s retrieval-based systems "computers will become generative, context-aware, and always running continuously"

Jensen Huang says AI will need 1,000x the computing energy we have today, and that estimate might still be too low Thats's because AI will not work like today’s retrieval-based systems "computers will become generative, context-aware, and always running continuously"

Haider.

22,288 次观看 • 1 个月前

Jensen Huang congratulates OpenAI on Stargate Norway. It will run on GB300 Superchips, scaling to hundreds of thousands of GPUs - purpose-built for training, reasoning, and real-time inference. “Just as electricity and the internet became foundational to modern life, AI will become essential infrastructure. Every country will build it. Every industry will depend on it.”

Jensen Huang congratulates OpenAI on Stargate Norway. It will run on GB300 Superchips, scaling to hundreds of thousands of GPUs - purpose-built for training, reasoning, and real-time inference. “Just as electricity and the internet became foundational to modern life, AI will become essential infrastructure. Every country will build it. Every industry will depend on it.”

vitrupo

73,644 次观看 • 10 个月前

AI Agents will completely transform how we interact with corporate knowledge by applying reasoning across multiple data sources. Here's an example of ChatGPT's new knowledge feature to pull information from Box and Gmail to combine it an answer a question.

AI Agents will completely transform how we interact with corporate knowledge by applying reasoning across multiple data sources. Here's an example of ChatGPT's new knowledge feature to pull information from Box and Gmail to combine it an answer a question.

Aaron Levie

73,669 次观看 • 7 个月前

Fei-Fei Li says real AI needs eyes. Not to look, but to move in 3D space. Spatial intelligence gave animals agency, and machines will need it too. And once AI can act in space, we can simulate limitless environments -- to train, explore, and create alongside them.

Fei-Fei Li says real AI needs eyes. Not to look, but to move in 3D space. Spatial intelligence gave animals agency, and machines will need it too. And once AI can act in space, we can simulate limitless environments -- to train, explore, and create alongside them.

vitrupo

171,579 次观看 • 1 年前