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。

"World models or cognitive models is absolutely fundamental. Having the ability to generalize abstract knowledge is fundamental." - Gary Marcus, neuro-symbolic AI expert & cognitive scientist, at the DKGcon 2024, at the DKGcon 2024

OriginTrail

103,687 subscribers

26,811 次观看 • 1 年前 •via X (Twitter)

科学技术教育

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

"There may be some neuro-symbolic hybrid much better than we have right now that smoothly transitions between data in a corpus and abstract representations. The kinds of stuff you guys are doing are compatible with that." — Gary Marcus ⏱️10 highlights: 00:00 Intro: Gary Marcus & Brana Rakic at DKGCon 02:22 Why AI took a wrong turn 04:42 Deep learning vs. deep understanding 06:16 World models: a foundational idea 08:12 The building blocks of trustworthy AI 09:44 Neuro-symbolic AI & abstract relationships 13:04 Safety is an ecosystem - the airplane lesson 15:03 The real risk of AI agents today 18:36 Testing AI like we test new drugs 22:49 Where OpenAI is headed The signal is clear: the path forward runs through world models, abstraction, and an open, accountable ecosystem.

"There may be some neuro-symbolic hybrid much better than we have right now that smoothly transitions between data in a corpus and abstract representations. The kinds of stuff you guys are doing are compatible with that." — Gary Marcus ⏱️10 highlights: 00:00 Intro: Gary Marcus & Brana Rakic at DKGCon 02:22 Why AI took a wrong turn 04:42 Deep learning vs. deep understanding 06:16 World models: a foundational idea 08:12 The building blocks of trustworthy AI 09:44 Neuro-symbolic AI & abstract relationships 13:04 Safety is an ecosystem - the airplane lesson 15:03 The real risk of AI agents today 18:36 Testing AI like we test new drugs 22:49 Where OpenAI is headed The signal is clear: the path forward runs through world models, abstraction, and an open, accountable ecosystem.

OriginTrail

227,307 次观看 • 1 个月前

The UK AISI might be the most situationally aware government entity in the world today. Today on The Cognitive Revolution Podcast, Chief Scientist Geoffrey Irving surveys the AI landscape. While jailbreaking frontier models is getting harder, the AISI Red Team has never failed. 👀

The UK AISI might be the most situationally aware government entity in the world today. Today on The Cognitive Revolution Podcast, Chief Scientist Geoffrey Irving surveys the AI landscape. While jailbreaking frontier models is getting harder, the AISI Red Team has never failed. 👀

Nathan Labenz

21,507 次观看 • 4 个月前

2024 Batch Reporting to Lokesh Sir, strong decision making is the fundamental quality of a leader at top🙏

2024 Batch Reporting to Lokesh Sir, strong decision making is the fundamental quality of a leader at top🙏

Chakradhar Rajaneni

58,479 次观看 • 2 年前

Andrej Karpathy: Internet training data is terrible, so big models end up compressing "memory" instead of doing cognitive work Use intelligent models to filter to the cognitive core With cleaner data, smaller models, likely distilled from a stronger one, are enough

Andrej Karpathy: Internet training data is terrible, so big models end up compressing "memory" instead of doing cognitive work Use intelligent models to filter to the cognitive core With cleaner data, smaller models, likely distilled from a stronger one, are enough

Haider.

384,224 次观看 • 9 个月前

Relive the finest moment in English men's football with ‘Looking back at ... The 1966 World Cup'. This cognitive book is suitable for people living with dementia or other cognitive difficulties, so everyone can enjoy. Buy now at @cognitivebook England

Relive the finest moment in English men's football with ‘Looking back at ... The 1966 World Cup'. This cognitive book is suitable for people living with dementia or other cognitive difficulties, so everyone can enjoy. Buy now at @cognitivebook England

Alzheimer's Society

55,589 次观看 • 1 年前

OpenAI’s GPT-5 was supposed to be a victory but instead it’s a disappointment, says Gary Marcus for CNBC, while championing a promising path to advance AI: “I think this really is the moment for neuro-symbolic AI.”

OpenAI’s GPT-5 was supposed to be a victory but instead it’s a disappointment, says Gary Marcus for CNBC, while championing a promising path to advance AI: “I think this really is the moment for neuro-symbolic AI.”

OriginTrail

33,090 次观看 • 11 个月前

Anthropic's CEO Dario Amodei: There are only a small number of years before AI models surpass the cognitive capabilities of most humans at most things. A set of AI agents more capable than most humans at most things, coordinating at superhuman speed. That level of capability is something the world has never seen before.

Anthropic's CEO Dario Amodei: There are only a small number of years before AI models surpass the cognitive capabilities of most humans at most things. A set of AI agents more capable than most humans at most things, coordinating at superhuman speed. That level of capability is something the world has never seen before.

Open Source Intel

97,853 次观看 • 5 个月前

🗣️”OriginTrail technology is built in such a way that it's privacy preserving and it's completely open-source, so you can see in-depth what exactly is happening and that your data is not leaving your device." ⦁ Brana Rakic, CTO of OriginTrail at the DKGcon 2024

🗣️”OriginTrail technology is built in such a way that it's privacy preserving and it's completely open-source, so you can see in-depth what exactly is happening and that your data is not leaving your device." ⦁ Brana Rakic, CTO of OriginTrail at the DKGcon 2024

OriginTrail

38,059 次观看 • 9 个月前

Alexandr Wang says "it's a strange time to be a scientist, but an exciting time for science." AI models develop intuitions humans cannot possess. Scientists become interpreters of those discoveries. A fundamental role reversal in how the frontier of knowledge expands.

Alexandr Wang says "it's a strange time to be a scientist, but an exciting time for science." AI models develop intuitions humans cannot possess. Scientists become interpreters of those discoveries. A fundamental role reversal in how the frontier of knowledge expands.

vitrupo

125,640 次观看 • 1 年前

“LLMs are polluting the Internet with hallucinations & convincing problematic content which naturally comes from centralized sources, introducing a lot of bias.” - Brana Rakic, Sub0 2024 🎯 Highlights: 00:35 AI threats: misinformation & deepfakes 03:08 Left-right brain & neuro-symbolic AI 06:05 The neuro-symbolic AI stack 13:06 What is dRAG? 23:30 Real-world enterprise adoption Get ready for sub0 2025, where Branimir returns to the stage as a speaker!

“LLMs are polluting the Internet with hallucinations & convincing problematic content which naturally comes from centralized sources, introducing a lot of bias.” - Brana Rakic, Sub0 2024 🎯 Highlights: 00:35 AI threats: misinformation & deepfakes 03:08 Left-right brain & neuro-symbolic AI 06:05 The neuro-symbolic AI stack 13:06 What is dRAG? 23:30 Real-world enterprise adoption Get ready for sub0 2025, where Branimir returns to the stage as a speaker!

OriginTrail

26,702 次观看 • 9 个月前

Ethernet pioneer Dr. Bob Metcalfe joined Brana Rakic at DKGcon 2025 to discuss how blockchains, LLMs and the DKG converge to power trusted intelligence networks. 00:00 – AI & knowledge graph convergence 04:35 – Internet pathologies 13:40 – Decentralized, user-defined filters 23:10 – Why interoperability matters

Ethernet pioneer Dr. Bob Metcalfe joined Brana Rakic at DKGcon 2025 to discuss how blockchains, LLMs and the DKG converge to power trusted intelligence networks. 00:00 – AI & knowledge graph convergence 04:35 – Internet pathologies 13:40 – Decentralized, user-defined filters 23:10 – Why interoperability matters

OriginTrail

12,634 次观看 • 7 个月前

being broke in ur 20s but good at vibe coding and intelligent with infinite energy and cognitive ability with max brain elasticity is not the same as being broke in ur 30s let alone broke in ur 40s with declining energy, cognitive ability, adaptability

being broke in ur 20s but good at vibe coding and intelligent with infinite energy and cognitive ability with max brain elasticity is not the same as being broke in ur 30s let alone broke in ur 40s with declining energy, cognitive ability, adaptability

gainzy

56,722 次观看 • 5 个月前

Ray Kurzweil says AI models will exceed human verbal ability in a year or two and the next step is large event models that enable robots to perform physical tasks

Ray Kurzweil says AI models will exceed human verbal ability in a year or two and the next step is large event models that enable robots to perform physical tasks

Tsarathustra

71,501 次观看 • 2 年前

The IG models not being honest about their work with having the ability to influence is a form of grooming.

The IG models not being honest about their work with having the ability to influence is a form of grooming.

Mageba

81,917 次观看 • 8 个月前

Here's a longer version of Walz “expansion of Israel” comment, with a bit more context before and after: “Iran, or I– uh– Israel's ability to be able to defend itself is absolutely fundamental. Getting its hostages back: fundamental. And ending the humanitarian crisis in Gaza. But the expansion of Israel and its proxies is an absolute, fundamental necessity for the United States [to] have the steady leadership there. You saw it experienced today, where, along with our Israeli partners and our coalition, able to stop the incoming attack. But what's fundamental here is that steady leadership is gonna to matter.”

Here's a longer version of Walz “expansion of Israel” comment, with a bit more context before and after: “Iran, or I– uh– Israel's ability to be able to defend itself is absolutely fundamental. Getting its hostages back: fundamental. And ending the humanitarian crisis in Gaza. But the expansion of Israel and its proxies is an absolute, fundamental necessity for the United States [to] have the steady leadership there. You saw it experienced today, where, along with our Israeli partners and our coalition, able to stop the incoming attack. But what's fundamental here is that steady leadership is gonna to matter.”

Decensored News

16,799 次观看 • 1 年前

Anthropic’s CEO says AI-driven robotics could become a multi-trillion-dollar industry as models learn to generalize in the real world. Robotics is where AI starts to touch the $50T+ physical economy across manufacturing, logistics, agriculture, construction and healthcare.

Anthropic’s CEO says AI-driven robotics could become a multi-trillion-dollar industry as models learn to generalize in the real world. Robotics is where AI starts to touch the $50T+ physical economy across manufacturing, logistics, agriculture, construction and healthcare.

Shay Boloor

73,843 次观看 • 5 个月前

Jasmine Crockett floats removing Trump from office via 25A "We need to talk about whether or not he really has the cognitive ability because he is the current sitting president."

Jasmine Crockett floats removing Trump from office via 25A "We need to talk about whether or not he really has the cognitive ability because he is the current sitting president."

Breitbart News

494,236 次观看 • 1 年前

Professor Elan Barenholtz, cognitive scientist at Florida Atlantic University, joins TOE to discuss one of the most unsettling ideas in cognitive science: that language is a self-contained, autoregressive system with no inherent connection to the external world. In this mind-altering episode, he explains why AI’s mastery of language without meaning forces us to rethink the nature of mind, perception, and reality itself... full episode🔗below

Professor Elan Barenholtz, cognitive scientist at Florida Atlantic University, joins TOE to discuss one of the most unsettling ideas in cognitive science: that language is a self-contained, autoregressive system with no inherent connection to the external world. In this mind-altering episode, he explains why AI’s mastery of language without meaning forces us to rethink the nature of mind, perception, and reality itself... full episode🔗below

Curt Jaimungal

111,788 次观看 • 1 年前