Video yükleniyor...

Video Yüklenemedi

Bu video yüklenirken bir sorun oluştu. Bu geçici bir ağ sorunundan kaynaklanıyor olabilir veya video kullanılamıyor olabilir.

Ana Sayfaya Dön

Supercomputers process huge amounts of data very quickly, and this power can be used to train AI models that advance research in cancer diagnoses, develop new medicines & sustainable materials. Watch Science Secretary Peter Kyle switch on the UK’s most powerful supercomputer👀

Department for Science, Innovation and Technology

214,127 subscribers

26,362 görüntüleme • 1 yıl önce •via X (Twitter)

Bilim & Teknoloji

Anya Rossi• Live Now

Private livecam show

0 Yorum

Yorum bulunmuyor

Orijinal gönderinin yorumları burada görünecek

Benzer Videolar

🚀 The UK’s most powerful AI supercomputer, Isambard-AI, is now live in Bristol. Officially launched at the NCC - Innovating for Industry today by Peter Kyle, from Department for Science, Innovation and Technology. It can complete 80 years of human work in one second. A new era for research begins🔬 Read more

🚀 The UK’s most powerful AI supercomputer, Isambard-AI, is now live in Bristol. Officially launched at the NCC - Innovating for Industry today by Peter Kyle, from Department for Science, Innovation and Technology. It can complete 80 years of human work in one second. A new era for research begins🔬 Read more

University of Bristol

12,534 görüntüleme • 1 yıl önce

Phase 1 of Isambard-AI, which will be the UK’s most powerful supercomputer, has officially been switched on🎉 At top 2 of the ISC Green500 list, we’re home to the most sustainable supercomputer in the UK and the 2nd most efficient in existence♻ Department for Science, Innovation and Technology HPE @NVIDIA

Phase 1 of Isambard-AI, which will be the UK’s most powerful supercomputer, has officially been switched on🎉 At top 2 of the ISC Green500 list, we’re home to the most sustainable supercomputer in the UK and the 2nd most efficient in existence♻ Department for Science, Innovation and Technology HPE @NVIDIA

University of Bristol

16,960 görüntüleme • 2 yıl önce

MD Anderson’s newly launched Institute for Data Science in Oncology combines our unparalleled scientific and clinical expertise with the power of data science to address cancer patients’ most urgent needs. Watch how we’re using data to advance discoveries. #EndCancer

MD Anderson’s newly launched Institute for Data Science in Oncology combines our unparalleled scientific and clinical expertise with the power of data science to address cancer patients’ most urgent needs. Watch how we’re using data to advance discoveries. #EndCancer

UT MD Anderson

10,451 görüntüleme • 2 yıl önce

🇺🇸 NVIDIA CEO HUANG: AMERICA IS BUILDING THE FUTURE OF SCIENCE WITH AI "Today, we're announcing that the Department of Energy is partnering with NVIDIA to build seven new AI supercomputers to advance our nation's science. Computing is the fundamental instrument of science. Every future supercomputer will be GPU based. Physics simulation will not go away, but it can be augmented with AI models working together. Classical computing could be enhanced to understand nature using quantum computing. These laboratories are impossible to experiment at the scale and speed we need to unless they're robotic laboratories." Source: amit

🇺🇸 NVIDIA CEO HUANG: AMERICA IS BUILDING THE FUTURE OF SCIENCE WITH AI "Today, we're announcing that the Department of Energy is partnering with NVIDIA to build seven new AI supercomputers to advance our nation's science. Computing is the fundamental instrument of science. Every future supercomputer will be GPU based. Physics simulation will not go away, but it can be augmented with AI models working together. Classical computing could be enhanced to understand nature using quantum computing. These laboratories are impossible to experiment at the scale and speed we need to unless they're robotic laboratories." Source: amit

Mario Nawfal

53,173 görüntüleme • 9 ay önce

#NewPaper The first microscope, invented in the 16th century, was designed to unlock the secrets of the microscopic world. Today, as many fields become increasingly data-driven, there is a pressing need for new types of microscopes---tools that help us zoom in, explore, and understand complex data. We call these tools "algorithmic microscopes." Introducing the Vendiscope: The first algorithmic microscope for data collections. 🔬 The Vendiscope maximizes the probability-weighted Vendi Score of a dataset to assign a weight to each element in the collection. This weight represents a data point's contribution to the overall diversity of the collection. These weights enable high-resolution data analysis at scale. We use them to zoom in on datasets across three domains: biology, materials science, & AI. 🧬 Biology: We used the Vendiscope on the protein universe, which contains nearly 250 million proteins. We found that nearly 200 million of the proteins are near-duplicates of each other and that AlphaFold fails on proteins that contribute most to the diversity of the protein universe. (See GIF below). 🪜 Materials Science: We used the Vendiscope on the Materials Project database, which contains 170K materials as of today. We found that 85% of crystals with formation energy data are near-duplicates of each other and that ML models for materials property prediction struggle with materials that contribute most to diversity. 🤖 Artificial Intelligence: We applied the Vendiscope to CIFAR-10, a benchmark dataset containing 50K images. We found duplicates. We applied the Vendiscope to analyze state-of-the-art generative models trained on this dataset. We found the best generative models memorize training data, as is known in the AI literature. However, we can do more with the Vendiscope and characterize the type of samples that get memorized. We found that data points contributing least to diversity are more prone to memorization by these generative models. 🧠 "Our findings demonstrate that the Vendiscope can serve as a powerful tool for data-driven science, providing a systematic and scalable way to identify duplicates and outliers, as well as pinpointing samples prone to memorization and those that models may struggle to predict---even before training." 💫 "The Vendiscope provides a unified framework for analyzing complex data at scale. Researchers, engineers, and data auditors can use the Vendiscope to audit datasets, identify potential biases, and refine data collection practices. For AI ethicists, the Vendiscope offers a critical lens to understand how models interact with data, particularly in the context of bias, memorization, and data fairness, enabling better mitigation strategies to prevent undesirable outcomes in AI deployment. For scientists, the Vendiscope represents a new companion in the discovery process." #VendiScoring #AlgorithmicMicroscopy Link to paper: Authors: Amey Pasarkar (Amey Pasarkar) and Adji Bousso Dieng (@adjiboussodieng)

#NewPaper The first microscope, invented in the 16th century, was designed to unlock the secrets of the microscopic world. Today, as many fields become increasingly data-driven, there is a pressing need for new types of microscopes---tools that help us zoom in, explore, and understand complex data. We call these tools "algorithmic microscopes." Introducing the Vendiscope: The first algorithmic microscope for data collections. 🔬 The Vendiscope maximizes the probability-weighted Vendi Score of a dataset to assign a weight to each element in the collection. This weight represents a data point's contribution to the overall diversity of the collection. These weights enable high-resolution data analysis at scale. We use them to zoom in on datasets across three domains: biology, materials science, & AI. 🧬 Biology: We used the Vendiscope on the protein universe, which contains nearly 250 million proteins. We found that nearly 200 million of the proteins are near-duplicates of each other and that AlphaFold fails on proteins that contribute most to the diversity of the protein universe. (See GIF below). 🪜 Materials Science: We used the Vendiscope on the Materials Project database, which contains 170K materials as of today. We found that 85% of crystals with formation energy data are near-duplicates of each other and that ML models for materials property prediction struggle with materials that contribute most to diversity. 🤖 Artificial Intelligence: We applied the Vendiscope to CIFAR-10, a benchmark dataset containing 50K images. We found duplicates. We applied the Vendiscope to analyze state-of-the-art generative models trained on this dataset. We found the best generative models memorize training data, as is known in the AI literature. However, we can do more with the Vendiscope and characterize the type of samples that get memorized. We found that data points contributing least to diversity are more prone to memorization by these generative models. 🧠 "Our findings demonstrate that the Vendiscope can serve as a powerful tool for data-driven science, providing a systematic and scalable way to identify duplicates and outliers, as well as pinpointing samples prone to memorization and those that models may struggle to predict---even before training." 💫 "The Vendiscope provides a unified framework for analyzing complex data at scale. Researchers, engineers, and data auditors can use the Vendiscope to audit datasets, identify potential biases, and refine data collection practices. For AI ethicists, the Vendiscope offers a critical lens to understand how models interact with data, particularly in the context of bias, memorization, and data fairness, enabling better mitigation strategies to prevent undesirable outcomes in AI deployment. For scientists, the Vendiscope represents a new companion in the discovery process." #VendiScoring #AlgorithmicMicroscopy Link to paper: Authors: Amey Pasarkar (Amey Pasarkar) and Adji Bousso Dieng (@adjiboussodieng)

Vertaix® (AI & Science)

34,762 görüntüleme • 1 yıl önce

Palantir Technologies is one of the most secretive and powerful companies in Silicon Valley. The company develops software that can quickly analyze massive amounts of data, used by government agencies, the military and major corporations. Ali Rogin speaks with Michael Steinberger, author of the new book “The Philosopher in the Valley: Alex Karp, Palantir and the Rise of the Surveillance State,” for more.

Palantir Technologies is one of the most secretive and powerful companies in Silicon Valley. The company develops software that can quickly analyze massive amounts of data, used by government agencies, the military and major corporations. Ali Rogin speaks with Michael Steinberger, author of the new book “The Philosopher in the Valley: Alex Karp, Palantir and the Rise of the Surveillance State,” for more.

PBS News

20,613 görüntüleme • 7 ay önce

SpaceX COO Gwynne Shotwell speaking about xAI's plans: • xAI will build 1.2 gigawatts of power for its supercomputer now and for all future data centers • Expanding the biggest Tesla Megapack system in the world — enough backup power for all of Memphis and more than Southaven needs (where the data center is) • New water recycling plants to save about 4.7 billion gallons of water each year from the Memphis aquifer • Thousands of new jobs mostly for people in Memphis, plus cheaper electricity for Americans • Plans to put supercomputers in space (orbit) to use huge amounts of energy more efficiently

SpaceX COO Gwynne Shotwell speaking about xAI's plans: • xAI will build 1.2 gigawatts of power for its supercomputer now and for all future data centers • Expanding the biggest Tesla Megapack system in the world — enough backup power for all of Memphis and more than Southaven needs (where the data center is) • New water recycling plants to save about 4.7 billion gallons of water each year from the Memphis aquifer • Thousands of new jobs mostly for people in Memphis, plus cheaper electricity for Americans • Plans to put supercomputers in space (orbit) to use huge amounts of energy more efficiently

Dima Zeniuk

433,942 görüntüleme • 5 ay önce

Training robot foundation models faces two key hurdles: how to get enough data to train an effective model, and how to make sure that new skills can be acquired quickly. The team at Rhoda AI believes that the answer is training Direct Video Action models from web data. Web data is plentiful, to the point where Rhoda can train their base model on hundreds of years of video data. And then, with the addition of robot data, they can quickly adapt it to new tasks with as little as 20 hours of in-domain data, performing complex, multi-step manipulation tasks with their purpose-built video foundation model. Tongzhou Mu 🤖🦾🦿 Eric Chan and Changan Chen joined us to talk more about their approach. Watch Episode #79 of RoboPapers, with Michael Cho - Rbt/Acc, Chris Paxton, and Jiafei Duan, to learn more!

Training robot foundation models faces two key hurdles: how to get enough data to train an effective model, and how to make sure that new skills can be acquired quickly. The team at Rhoda AI believes that the answer is training Direct Video Action models from web data. Web data is plentiful, to the point where Rhoda can train their base model on hundreds of years of video data. And then, with the addition of robot data, they can quickly adapt it to new tasks with as little as 20 hours of in-domain data, performing complex, multi-step manipulation tasks with their purpose-built video foundation model. Tongzhou Mu 🤖🦾🦿 Eric Chan and Changan Chen joined us to talk more about their approach. Watch Episode #79 of RoboPapers, with Michael Cho - Rbt/Acc, Chris Paxton, and Jiafei Duan, to learn more!

RoboPapers

24,475 görüntüleme • 3 ay önce

“We’ve been talking about breakthrough innovation for decades—AI may finally be the key. It gives us the ability to process extraordinary amounts of data and unlock entirely new ideas.” Rep. Brad Schneider on why he’s most excited about AI’s potential in health care. 📺 Watch the full event: Event sponsored by Takeda #TheHillHealth #healthTech #AI

“We’ve been talking about breakthrough innovation for decades—AI may finally be the key. It gives us the ability to process extraordinary amounts of data and unlock entirely new ideas.” Rep. Brad Schneider on why he’s most excited about AI’s potential in health care. 📺 Watch the full event: Event sponsored by Takeda #TheHillHealth #healthTech #AI

The Hill Events

17,211 görüntüleme • 10 ay önce

"It all comes down to compute." Dmitri Alperovitch from Silverado Policy Accelerator explains: "The single most important input to winning is compute the processing power used to train and run AI models. Let me say that plainly, because it is defining that everything that congress and this committee do on AI policy. The binding constraint in this competition is not talent, it is not data, and it is not cash. Both the United States and China have talented researchers and financial resources, and we have learned that frontier models trained on generic data routinely outperform smaller models, trained on more specialized inputs. It all comes down to compute."

"It all comes down to compute." Dmitri Alperovitch from Silverado Policy Accelerator explains: "The single most important input to winning is compute the processing power used to train and run AI models. Let me say that plainly, because it is defining that everything that congress and this committee do on AI policy. The binding constraint in this competition is not talent, it is not data, and it is not cash. Both the United States and China have talented researchers and financial resources, and we have learned that frontier models trained on generic data routinely outperform smaller models, trained on more specialized inputs. It all comes down to compute."

Select Committee on China

41,578 görüntüleme • 4 ay önce

Amjad Masad says true AGI means an AI that can enter new situations and quickly learn how to achieve goals Even the latest models still struggle with tasks outside their training data. The real breakthrough will come when AI can train itself -- like AlphaGo playing billions of games against itself

Amjad Masad says true AGI means an AI that can enter new situations and quickly learn how to achieve goals Even the latest models still struggle with tasks outside their training data. The real breakthrough will come when AI can train itself -- like AlphaGo playing billions of games against itself

Haider.

163,410 görüntüleme • 1 yıl önce

Cats may help in the fight against cancer in humans Scientists analyzed nearly 500 samples from 13 types of cancer in domestic cats and decoded their genomes. They compared tumor tissue with healthy tissue and found mutations that match those seen in humans — including in the TP53 gene, one of the key genes involved in cancer development. This matters because cats live in the same environment as people and develop similar diseases. Their genome can be used to test new treatment methods — first in veterinary medicine and later in human medicine. This approach is known as “One Medicine,” where science benefits both animals and humans. Research into feline cancer could accelerate the search for new therapies and help scientists better understand how tumors develop in humans. The video is for illustrative purposes.

Cats may help in the fight against cancer in humans Scientists analyzed nearly 500 samples from 13 types of cancer in domestic cats and decoded their genomes. They compared tumor tissue with healthy tissue and found mutations that match those seen in humans — including in the TP53 gene, one of the key genes involved in cancer development. This matters because cats live in the same environment as people and develop similar diseases. Their genome can be used to test new treatment methods — first in veterinary medicine and later in human medicine. This approach is known as “One Medicine,” where science benefits both animals and humans. Research into feline cancer could accelerate the search for new therapies and help scientists better understand how tumors develop in humans. The video is for illustrative purposes.

NEXTA

40,255 görüntüleme • 5 ay önce

Today, we’re introducing Forge, a system for enterprises to build frontier-grade AI models grounded in their proprietary knowledge. 🌎 Forge bridges the gap between generic AI and enterprise-specific needs. Instead of relying on broad, public data, organizations can train models that understand their internal context embedded within systems, workflows, and policies, aligning AI with their unique operations. We have already partnered with world-leading organizations, like ASML, DSO National Laboratories Singapore, Ericsson, European Space Agency, Home Team Science and Technology Agency (HTX) Singapore and Reply to train models on the proprietary data that powers their most complex systems and future-defining technologies.

Today, we’re introducing Forge, a system for enterprises to build frontier-grade AI models grounded in their proprietary knowledge. 🌎 Forge bridges the gap between generic AI and enterprise-specific needs. Instead of relying on broad, public data, organizations can train models that understand their internal context embedded within systems, workflows, and policies, aligning AI with their unique operations. We have already partnered with world-leading organizations, like ASML, DSO National Laboratories Singapore, Ericsson, European Space Agency, Home Team Science and Technology Agency (HTX) Singapore and Reply to train models on the proprietary data that powers their most complex systems and future-defining technologies.

Mistral AI

441,460 görüntüleme • 4 ay önce

.Secretary Marco Rubio: "We remain the world’s sole global superpower—the most powerful country on earth. We have the largest economy, we have the most formidable and powerful military ever known to man, by the way. Our dollar remains that reserve currency of the world. Our language is used extensively throughout the world... but all of this means very little if that power is not used to protect the people who built it."

.Secretary Marco Rubio: "We remain the world’s sole global superpower—the most powerful country on earth. We have the largest economy, we have the most formidable and powerful military ever known to man, by the way. Our dollar remains that reserve currency of the world. Our language is used extensively throughout the world... but all of this means very little if that power is not used to protect the people who built it."

Rapid Response 47

368,137 görüntüleme • 2 ay önce

🚨 WEF - “The Future is Ai” “$600 Billion on Data Centres” Larry Fink - “it will be bigger” They are building data centres in huge numbers across the UK & Europe as well - ask yourself - why? Data Centres consume huge amounts of - hey Grok how many homes could be fed with water and energy - equal to that of one large data centre?

🚨 WEF - “The Future is Ai” “$600 Billion on Data Centres” Larry Fink - “it will be bigger” They are building data centres in huge numbers across the UK & Europe as well - ask yourself - why? Data Centres consume huge amounts of - hey Grok how many homes could be fed with water and energy - equal to that of one large data centre?

Concerned Citizen

34,935 görüntüleme • 6 ay önce

🚨 WOW! HHS SEC. BOBBY KENNEDY just DROPPED this line on curing cancer: "I just was reading about a dog that had cancer and his owner used AI to develop a personalized treatment that cured the cancer! We’re going to see that now across medicine. And we're very excited about it." "AI is potentially dangerous, but has the capacity to bring great things to humanity." "AI may, at some day, make FDA irrelevant. It's gonna give us the capacity to develop new drugs, personalized medicine for EVERY citizen."

🚨 WOW! HHS SEC. BOBBY KENNEDY just DROPPED this line on curing cancer: "I just was reading about a dog that had cancer and his owner used AI to develop a personalized treatment that cured the cancer! We’re going to see that now across medicine. And we're very excited about it." "AI is potentially dangerous, but has the capacity to bring great things to humanity." "AI may, at some day, make FDA irrelevant. It's gonna give us the capacity to develop new drugs, personalized medicine for EVERY citizen."

Eric Daugherty

147,121 görüntüleme • 3 ay önce

I used NotebookLM to study Google new breakthrough with TurboQuant and used Video overview to study the subject, best learning tool in the world at the moment. TurboQuant: Redefining AI Efficiency with Extreme Compression Google Research has introduced TurboQuant, a suite of advanced algorithms designed to dramatically compress the data used by large language models and search engines. By utilizing specialized techniques like PolarQuant and Quantized Johnson-Lindenstrauss, the system transforms complex information into a compact "shorthand" that requires significantly less memory. This innovation specifically addresses the key-value cache bottleneck, allowing AI models to process massive amounts of text faster without losing accuracy. Testing demonstrates that these methods can shrink data size by six times while actually increasing operational speed on high-end hardware. Ultimately, these theoretical advancements enable more efficient semantic search and high-performance AI applications at a global scale.

I used NotebookLM to study Google new breakthrough with TurboQuant and used Video overview to study the subject, best learning tool in the world at the moment. TurboQuant: Redefining AI Efficiency with Extreme Compression Google Research has introduced TurboQuant, a suite of advanced algorithms designed to dramatically compress the data used by large language models and search engines. By utilizing specialized techniques like PolarQuant and Quantized Johnson-Lindenstrauss, the system transforms complex information into a compact "shorthand" that requires significantly less memory. This innovation specifically addresses the key-value cache bottleneck, allowing AI models to process massive amounts of text faster without losing accuracy. Testing demonstrates that these methods can shrink data size by six times while actually increasing operational speed on high-end hardware. Ultimately, these theoretical advancements enable more efficient semantic search and high-performance AI applications at a global scale.

Emily

14,648 görüntüleme • 4 ay önce