正在加载视频...

视频加载失败

返回首页

NEW! The Edge browser now has a "Learning Toolkit" mode 🎓 Make it easy to find #EdgeEDU features! Just type "edge://learning" ➗ Math Solver 📌 Citations 📃 PDF reader 📖 Immersive Reader 🔊 Read Aloud More details: #edtech #MIEExpert #MicrosoftEDU

mtholfsen's profile picture

Mike Tholfsen

65,719 subscribers

18,721 次观看 • 3 年前 •via X (Twitter)

教育#EdgeEDU#edtech#MIEExpert#MicrosoftEDU
LIVE

Anya Rossi• Live Now

Private livecam show

4 条评论

TricksyBell❣ 的头像
TricksyBell❣3 年前

Nearly as good as edge://surf 🏄‍♀️ 🤣

Javier Boncompte G. 的头像
Javier Boncompte G.3 年前

Are there any coming improvements for ink in PDFs? Like multiple pens, styles, better settings like office, etc?

Ajay 的头像
Ajay3 年前

@MicrosoftEdge

Mike Tholfsen 的头像
Mike Tholfsen3 年前

@MrKavanagh Yup. Not sure if it has rolled to main or still in beta builds there

相关视频

Copilot can now control your browser 🔥 Microsoft has just updated Edge to make it fully agentic - Enable Copilot Mode - Turn on Actions in Edge Preview - Give the AI any goal to achieve You can either let it work in the background or watch it perform actions. And there are 3 modes available to make it more or less autonomous (light, balanced, and strict).
0:37

Copilot can now control your browser 🔥 Microsoft has just updated Edge to make it fully agentic - Enable Copilot Mode - Turn on Actions in Edge Preview - Give the AI any goal to achieve You can either let it work in the background or watch it perform actions. And there are 3 modes available to make it more or less autonomous (light, balanced, and strict).

Paul Couvert

52,823 次观看 • 9 个月前

Want your own custom AI helper? With upcoming updates from OptimAI, it's super simple! 🤖✨ Just type one sentence and the network does all the smart work behind the scenes: - Super-fast edge computing + private learning - Secure global storage - Real human feedback to make it better In 3 easy steps: ✅ Drop your prompt ✅ OptimAI finds/refines/trains everything ✅ Get your personal AI agent, ready to work & keep improving! Anyone can build powerful AI now The future of AI is here, and it's for everyone. 🚀
0:40

Want your own custom AI helper? With upcoming updates from OptimAI, it's super simple! 🤖✨ Just type one sentence and the network does all the smart work behind the scenes: - Super-fast edge computing + private learning - Secure global storage - Real human feedback to make it better In 3 easy steps: ✅ Drop your prompt ✅ OptimAI finds/refines/trains everything ✅ Get your personal AI agent, ready to work & keep improving! Anyone can build powerful AI now The future of AI is here, and it's for everyone. 🚀

OptimAI Network

15,714 次观看 • 4 个月前

I vibe coded a game called Aether Breach for my son today. Built the whole thing in about 2 hours using Cursor, Grok 4.5, and GPT Image 2.0. He gave it a try and really liked it. I just need to make Easy mode a bit more relaxed so he can make it to the later stages. The full game has DualSense support too. I published my WebHID DualSense npm package a few days ago, so adding controller support to new games is pretty straightforward now. Really happy with how it turned out.
1:23

I vibe coded a game called Aether Breach for my son today. Built the whole thing in about 2 hours using Cursor, Grok 4.5, and GPT Image 2.0. He gave it a try and really liked it. I just need to make Easy mode a bit more relaxed so he can make it to the later stages. The full game has DualSense support too. I published my WebHID DualSense npm package a few days ago, so adding controller support to new games is pretty straightforward now. Really happy with how it turned out.

AshutoshShrivastava

115,948 次观看 • 17 天前

TradingView to Screener & Marketsmith Extension ⚠️Please before asking in comment or DM, how to install, please read the full post, i have given the instructions also, and if you still cant follow it, You will find plenty of video on youtube on how to install the plugin, or ask AI⚠️ 🙏This was a killer timesaver tool for me, Hope it helps all, please share your feedbacks or gratitude in comments, will be sharing more plugin like this in future as well ☑️Overview This Chrome Extension natively integrates with TradingView, adding quick-access buttons to your top chart header. These buttons allow you to instantly view fundamental financial data for the currently active stock symbol on or evaluate it on MarketSmith India. > Features : > Screener New Tab: Opens the consolidated financials page for the current active TradingView symbol in a new browser tab. > Screener Splitscreen: Toggles a bottom-half split screen (iframe) inside the TradingView tab to load the page directly over your charts, letting you perform fundamental and technical analysis synchronously. > Marketsmith New Tab: Opens the stock evaluation page on MarketSmith India for the active symbol in a new tab. > This extension utilizes the Manifest V3 standard APIs and works smoothly across all modern Chromium-based browsers, including: Google Chrome Microsoft Edge Brave Browser Vivaldi Opera Arc Browser ⭐ Installation Instructions (Local/Unpacked) -> 1)Download or keep this entire extension folder on your computer. 1) Download or keep this entire extension folder on your computer: For Google Chrome: type chrome://extensions/ in the address bar. For Microsoft Edge: type edge://extensions/ in the address bar. For Brave: type brave://extensions/ in the address bar. Enable Developer mode using the toggle switch (usually located in the top-right corner). 3) Click the Load unpacked button that appears. 4) In the file explorer popup, choose this exact directory (the folder containing the manifest.json file). 5) The extension is now successfully installed! Open up TradingView to play around with the newly injected buttons in the top header. Github Repo ->
0:50

TradingView to Screener & Marketsmith Extension ⚠️Please before asking in comment or DM, how to install, please read the full post, i have given the instructions also, and if you still cant follow it, You will find plenty of video on youtube on how to install the plugin, or ask AI⚠️ 🙏This was a killer timesaver tool for me, Hope it helps all, please share your feedbacks or gratitude in comments, will be sharing more plugin like this in future as well ☑️Overview This Chrome Extension natively integrates with TradingView, adding quick-access buttons to your top chart header. These buttons allow you to instantly view fundamental financial data for the currently active stock symbol on or evaluate it on MarketSmith India. > Features : > Screener New Tab: Opens the consolidated financials page for the current active TradingView symbol in a new browser tab. > Screener Splitscreen: Toggles a bottom-half split screen (iframe) inside the TradingView tab to load the page directly over your charts, letting you perform fundamental and technical analysis synchronously. > Marketsmith New Tab: Opens the stock evaluation page on MarketSmith India for the active symbol in a new tab. > This extension utilizes the Manifest V3 standard APIs and works smoothly across all modern Chromium-based browsers, including: Google Chrome Microsoft Edge Brave Browser Vivaldi Opera Arc Browser ⭐ Installation Instructions (Local/Unpacked) -> 1)Download or keep this entire extension folder on your computer. 1) Download or keep this entire extension folder on your computer: For Google Chrome: type chrome://extensions/ in the address bar. For Microsoft Edge: type edge://extensions/ in the address bar. For Brave: type brave://extensions/ in the address bar. Enable Developer mode using the toggle switch (usually located in the top-right corner). 3) Click the Load unpacked button that appears. 4) In the file explorer popup, choose this exact directory (the folder containing the manifest.json file). 5) The extension is now successfully installed! Open up TradingView to play around with the newly injected buttons in the top header. Github Repo ->

Roshan Kumar

14,810 次观看 • 5 个月前

We are thrilled to introduce to you the latest version of the #SAFEMOONWALLET! Built to keep your experience on the cutting edge! 🙌 HIGHLIGHTS 🔵 dApp Browser Integration In our efforts to enhance your decentralized experience, users now have the option to utilize our dApp Browser! Seamlessly search and connect to decentralized applications without ever leaving the app. 🔵 Enhanced Navigation & Tools With streamlined access to key application features, enjoy swift navigation when searching for essential tools like the calculator, customer support, security features and more! 🔵 Built-in Web Browser Take advantage of surfing the web within the wallet and use the bookmark feature to ensure you’re always going to the right page! 🔵 Advanced Darker Mode To those who desire a more subdued interface, our new mode provides users the option to make dark mode, darker! 🔵 WalletConnect 2.0 Alignment Continued upgrades to keep pace with WalletConnect’s ever evolving tech. 🔵 Optimized Collection Viewing This performance improvement ensures you navigate your collectibles with ease and speed. 🔵 Application Refinements With our commitment to continual progression, we have implemented various other bug fixes and improvements. We hope you enjoy this update to the SafeMoon Wallet as much as we do! 🎉 INVENT | EXCITE | UNITE #SAFEMOON
0:43

We are thrilled to introduce to you the latest version of the #SAFEMOONWALLET! Built to keep your experience on the cutting edge! 🙌 HIGHLIGHTS 🔵 dApp Browser Integration In our efforts to enhance your decentralized experience, users now have the option to utilize our dApp Browser! Seamlessly search and connect to decentralized applications without ever leaving the app. 🔵 Enhanced Navigation & Tools With streamlined access to key application features, enjoy swift navigation when searching for essential tools like the calculator, customer support, security features and more! 🔵 Built-in Web Browser Take advantage of surfing the web within the wallet and use the bookmark feature to ensure you’re always going to the right page! 🔵 Advanced Darker Mode To those who desire a more subdued interface, our new mode provides users the option to make dark mode, darker! 🔵 WalletConnect 2.0 Alignment Continued upgrades to keep pace with WalletConnect’s ever evolving tech. 🔵 Optimized Collection Viewing This performance improvement ensures you navigate your collectibles with ease and speed. 🔵 Application Refinements With our commitment to continual progression, we have implemented various other bug fixes and improvements. We hope you enjoy this update to the SafeMoon Wallet as much as we do! 🎉 INVENT | EXCITE | UNITE #SAFEMOON

SafeMoon

162,434 次观看 • 3 年前

Geoffrey Hinton explains the coyote test for medical AI: radiology is already over the cliff, it just has not looked down yet "If you work as a radiologist, you're like the coyote that's already over the edge of the cliff, but hasn't yet looked down" "People should stop training radiologists now. It's just completely obvious that within 5 years deep learning is going to do better than radiologists" "It might be 10 years, but we got plenty of radiologists already" "There's going to be thousands of applications of the deep learning technology we currently have, especially stuff using fast chips" "Thousands of applications and hundreds of them are going to work, and those are the ones you're going to hear about" Most healthcare AI conversations still price the bottleneck as adoption. Hinton was pointing at a different bottleneck: medicine has specialties where the machine gets more examples than any human can. The brutal part is the error bar. If it is 5 years, training pipelines are already late. If it is 10, hospitals still have to redesign the workflow before the models arrive. - Geoffrey Hinton, Nobel laureate and pioneer of deep learning, at Creative Destruction Lab
0:50

Geoffrey Hinton explains the coyote test for medical AI: radiology is already over the cliff, it just has not looked down yet "If you work as a radiologist, you're like the coyote that's already over the edge of the cliff, but hasn't yet looked down" "People should stop training radiologists now. It's just completely obvious that within 5 years deep learning is going to do better than radiologists" "It might be 10 years, but we got plenty of radiologists already" "There's going to be thousands of applications of the deep learning technology we currently have, especially stuff using fast chips" "Thousands of applications and hundreds of them are going to work, and those are the ones you're going to hear about" Most healthcare AI conversations still price the bottleneck as adoption. Hinton was pointing at a different bottleneck: medicine has specialties where the machine gets more examples than any human can. The brutal part is the error bar. If it is 5 years, training pipelines are already late. If it is 10, hospitals still have to redesign the workflow before the models arrive. - Geoffrey Hinton, Nobel laureate and pioneer of deep learning, at Creative Destruction Lab

Karl Mehta

55,948 次观看 • 1 个月前

Experiments in progress. The one on the right has been learning for ~3 hours, the one in the middle for ~1 hour, and the one on the left just started a few minutes ago. The initial motivation for making the physical Atari was just to commit ourselves to a subset of algorithms that can make progress in this setup. This commitment rules out algorithms that require billions of samples to learn (or worse, require multiple environments running in parallel). Atari games are simple enough that we should be able to show learning on them in a short amount of time with no prior knowledge. Since then, I've realized that this setup is also a good way to compare different paradigms in robotics in a principled way. These paradigms are sim2real, learning from tele-operated data, and learning directly on the robots. So far, I have observed that getting sim2real to work reliably is hard. It requires tweaks that don't scale. Policies that can play perfectly in simulation fall apart because of latencies and the messiness of the real world. These aspects could be modeled to improve the simulation, but not without sinking significant human engineering hours. I have higher hopes for learning from tele-operated data, but that requires a human to learn the task first. These experiments are on my to-do list. I have to learn to play some of the games well through the robot. I’m half-decent at playing Pong and Ms Pacman now. Learning directly on robots is looking like the most promising approach. This approach takes away pesky distribution shifts and makes it possible to have algorithms that continually improve with more data and time without any human intervention. It feels great to let experiments run overnight and wake up to find improved policies. With learning on robots, I should, in principle, be able to go on a long vacation and come back to find better policies for complex tasks beyond Atari games. Whether that is possible with current learning algorithms is a different question.
1:05

Experiments in progress. The one on the right has been learning for ~3 hours, the one in the middle for ~1 hour, and the one on the left just started a few minutes ago. The initial motivation for making the physical Atari was just to commit ourselves to a subset of algorithms that can make progress in this setup. This commitment rules out algorithms that require billions of samples to learn (or worse, require multiple environments running in parallel). Atari games are simple enough that we should be able to show learning on them in a short amount of time with no prior knowledge. Since then, I've realized that this setup is also a good way to compare different paradigms in robotics in a principled way. These paradigms are sim2real, learning from tele-operated data, and learning directly on the robots. So far, I have observed that getting sim2real to work reliably is hard. It requires tweaks that don't scale. Policies that can play perfectly in simulation fall apart because of latencies and the messiness of the real world. These aspects could be modeled to improve the simulation, but not without sinking significant human engineering hours. I have higher hopes for learning from tele-operated data, but that requires a human to learn the task first. These experiments are on my to-do list. I have to learn to play some of the games well through the robot. I’m half-decent at playing Pong and Ms Pacman now. Learning directly on robots is looking like the most promising approach. This approach takes away pesky distribution shifts and makes it possible to have algorithms that continually improve with more data and time without any human intervention. It feels great to let experiments run overnight and wake up to find improved policies. With learning on robots, I should, in principle, be able to go on a long vacation and come back to find better policies for complex tasks beyond Atari games. Whether that is possible with current learning algorithms is a different question.

Khurram Javed

52,110 次观看 • 8 个月前

Just when I thought I couldn’t hate Monarch more. Guy announces a “Zero Edge” slot on Duel and a “Zero Edge” live blackjack game. You can’t just slap zero edge on something and call it revolutionary. It’s the same type of marketing buzzword as “provably fair,” which is already a running joke in this industry. Slapping ‘zero edge’ on a game doesn’t magically make it innovative. Without real transparency around in house server seeds that can be manipulated in the house’s favor, it’s hard for players to know what they’re actually getting. Nobody should touch them with a 10 foot pole. Probably fair is mostly a scam. There are some exceptions that actually implement true provably fair but most casinos slap the label on and have no idea what it means. I can’t believe this guy used to be one of my heroes. He’s blatantly launching garbage and he knows it. His team behind the scenes is trash and he admits it, and Duel has only gone downhill ever since Ossi got perma banned on Twitter. He was the only one carrying marketing and all things Duel on his back. Right now there’s no truly superior product in this space except Stake and honestly, that’s pretty disappointing.
0:29

Just when I thought I couldn’t hate Monarch more. Guy announces a “Zero Edge” slot on Duel and a “Zero Edge” live blackjack game. You can’t just slap zero edge on something and call it revolutionary. It’s the same type of marketing buzzword as “provably fair,” which is already a running joke in this industry. Slapping ‘zero edge’ on a game doesn’t magically make it innovative. Without real transparency around in house server seeds that can be manipulated in the house’s favor, it’s hard for players to know what they’re actually getting. Nobody should touch them with a 10 foot pole. Probably fair is mostly a scam. There are some exceptions that actually implement true provably fair but most casinos slap the label on and have no idea what it means. I can’t believe this guy used to be one of my heroes. He’s blatantly launching garbage and he knows it. His team behind the scenes is trash and he admits it, and Duel has only gone downhill ever since Ossi got perma banned on Twitter. He was the only one carrying marketing and all things Duel on his back. Right now there’s no truly superior product in this space except Stake and honestly, that’s pretty disappointing.

Yogi

30,968 次观看 • 4 个月前

Robots aren’t just the future — they’re already reshaping our world. At IIT Madras, the new in Robotics, offered under the Department of Engineering Design, is building the brains behind tomorrow’s intelligent machines. From motion planning and control to AI/ML and hands-on labs, this two-year program blends cutting-edge research with real-world industry relevance — powered by a strong robotics faculty team and a legacy of interdisciplinary innovation. Rooted in the success of the Inter-Disciplinary Dual Degree (IDDD) Robotics program, this two-year program brings together expert faculty members and a powerful blend of theory and hands-on learning. From robot design and analysis to motion planning, control systems, and AI/ML, students dive deep into the science and engineering behind next-generation robotic systems. Add immersive lab work and cutting-edge projects, and you have a curriculum that’s as practical as it is pioneering. Manufacturing, healthcare, logistics, and defence — robotics is reshaping them all. This program is about building the minds that will lead to that transformation. If the future runs on robots, it starts here. For more details visit: IIT Madras
2:52

Robots aren’t just the future — they’re already reshaping our world. At IIT Madras, the new in Robotics, offered under the Department of Engineering Design, is building the brains behind tomorrow’s intelligent machines. From motion planning and control to AI/ML and hands-on labs, this two-year program blends cutting-edge research with real-world industry relevance — powered by a strong robotics faculty team and a legacy of interdisciplinary innovation. Rooted in the success of the Inter-Disciplinary Dual Degree (IDDD) Robotics program, this two-year program brings together expert faculty members and a powerful blend of theory and hands-on learning. From robot design and analysis to motion planning, control systems, and AI/ML, students dive deep into the science and engineering behind next-generation robotic systems. Add immersive lab work and cutting-edge projects, and you have a curriculum that’s as practical as it is pioneering. Manufacturing, healthcare, logistics, and defence — robotics is reshaping them all. This program is about building the minds that will lead to that transformation. If the future runs on robots, it starts here. For more details visit: IIT Madras

IIT Madras

22,446 次观看 • 4 个月前

Why You Should Write 1) Putting words on paper freezes your thoughts, which gives your mind space to think deeper about a problem. 2) Writing lets you see if the epiphany you just had is brilliant or total bogus. 3) You want to attract smart people — because smart people read, and smart & successful people read the most. 4) For whatever field you're interested in, if you write about it well, the world's experts will reach out to you. 5) The higher you climb, the more likely you are to write frequently. Every serious executive I know is a voracious reader. Or, as former American president Harry Truman once said: “Not all readers are leaders but all leaders are readers.” 6) Writing accelerates your learning. If you want to understand a topic better, start teaching it. The act of putting ideas into your own words tattoos them into your mind. James Clear once wrote: "If you think you can learn a lot by reading a book, try writing one." 7) The investor Howard Marks once said to me: "In writing, as opposed to speaking, you can't BS. You can't BS in writing because the reader can easily go back and check your work. The logic has to be established. The explanation needs to be clear." 8) It frees up processing space: Mathematicians put their ideas on the whiteboard because jotting down their ideas lets them tackle harder problems. Writing is the same. By putting ideas onto the page, you can transcend the limits of scattered thinking and tackle more challenging problems.
0:40

Why You Should Write 1) Putting words on paper freezes your thoughts, which gives your mind space to think deeper about a problem. 2) Writing lets you see if the epiphany you just had is brilliant or total bogus. 3) You want to attract smart people — because smart people read, and smart & successful people read the most. 4) For whatever field you're interested in, if you write about it well, the world's experts will reach out to you. 5) The higher you climb, the more likely you are to write frequently. Every serious executive I know is a voracious reader. Or, as former American president Harry Truman once said: “Not all readers are leaders but all leaders are readers.” 6) Writing accelerates your learning. If you want to understand a topic better, start teaching it. The act of putting ideas into your own words tattoos them into your mind. James Clear once wrote: "If you think you can learn a lot by reading a book, try writing one." 7) The investor Howard Marks once said to me: "In writing, as opposed to speaking, you can't BS. You can't BS in writing because the reader can easily go back and check your work. The logic has to be established. The explanation needs to be clear." 8) It frees up processing space: Mathematicians put their ideas on the whiteboard because jotting down their ideas lets them tackle harder problems. Writing is the same. By putting ideas onto the page, you can transcend the limits of scattered thinking and tackle more challenging problems.

David Perell

99,861 次观看 • 2 年前

Dan Lanning shares why the best teams never stop growing. "You can't just grow in one area. It doesn't matter if you're a senior or a freshman or a coach that's been coaching 20 years - there's more to learn." "The game of football keeps changing. There's always people that are innovative. There's new adaptations. And it's the same thing in the game of life." Growth never stops. You have to make it a daily habit. The moment you think you've arrived - you've already fallen behind. "We wanna be a team that's seeking out information. Looking for opportunities to grow." "It's not growth in one field. It's growth in every field." The best teams are always learning. Always adapting. Always looking for an edge. You don't get to choose when the game changes. You only get to choose if you're ready when it does. (🎥George Wrighster Podcast )
0:49

Dan Lanning shares why the best teams never stop growing. "You can't just grow in one area. It doesn't matter if you're a senior or a freshman or a coach that's been coaching 20 years - there's more to learn." "The game of football keeps changing. There's always people that are innovative. There's new adaptations. And it's the same thing in the game of life." Growth never stops. You have to make it a daily habit. The moment you think you've arrived - you've already fallen behind. "We wanna be a team that's seeking out information. Looking for opportunities to grow." "It's not growth in one field. It's growth in every field." The best teams are always learning. Always adapting. Always looking for an edge. You don't get to choose when the game changes. You only get to choose if you're ready when it does. (🎥George Wrighster Podcast )

Coach AJ 🎯 Mental Fitness

13,174 次观看 • 1 个月前

I’m really proud of the results Wayve released today. 📊 For the first time we share quantitative results of our global generalisation journey. What did we learn? 💡 Over our first 8-weeks training in the US, our AI has rapidly adapted to its unique driving environment with just 500 hours worth of incremental US-specific training data, on top of our large scale foundation model's learning. What’s more, we just launched in Germany and first results are already 3x better — generalisation just keeps improving with new data. It is clear to me the future of autonomy will be unlocked by Embodied AI capable of generalising to new countries, vehicles and able to safely navigate challenging edge cases it has never seen before during training. Now we’re excited to kick things up a gear and work with the world’s leading automotive manufacturers to collect the data necessary to validate this technology and deploy it in consumer vehicles worldwide. Embodied AI is a generational opportunity and the pace of progress I'm seeing is incredible. We're building the strongest team in AI for automated driving. If you're excited about this mission, join us!
1:10

I’m really proud of the results Wayve released today. 📊 For the first time we share quantitative results of our global generalisation journey. What did we learn? 💡 Over our first 8-weeks training in the US, our AI has rapidly adapted to its unique driving environment with just 500 hours worth of incremental US-specific training data, on top of our large scale foundation model's learning. What’s more, we just launched in Germany and first results are already 3x better — generalisation just keeps improving with new data. It is clear to me the future of autonomy will be unlocked by Embodied AI capable of generalising to new countries, vehicles and able to safely navigate challenging edge cases it has never seen before during training. Now we’re excited to kick things up a gear and work with the world’s leading automotive manufacturers to collect the data necessary to validate this technology and deploy it in consumer vehicles worldwide. Embodied AI is a generational opportunity and the pace of progress I'm seeing is incredible. We're building the strongest team in AI for automated driving. If you're excited about this mission, join us!

Alex Kendall

21,933 次观看 • 1 年前

Excited to release a tool that makes the joy of live music creation more accessible 🎸 “Music FX DJ” is new and improved: Featuring a whole suite of new controls, sharing and remix features, and improved audio quality and diversity of outputs, to let you blend and warp genres and instruments in an ever evolving stream of music. It’s been an absolute pleasure to work with Jacob Collier over the past year on this. It was very affirming to find him pushing for the same things Google Magenta Project has always valued: making tools that intrinsically value creativity and agency over just imitation, embrace the “weird corners” of AI, encouraging veterans to express themselves in new ways, while empowering novices to express themselves musically maybe for the first time. Read details at: And in the 🧵 below…
0:56

Excited to release a tool that makes the joy of live music creation more accessible 🎸 “Music FX DJ” is new and improved: Featuring a whole suite of new controls, sharing and remix features, and improved audio quality and diversity of outputs, to let you blend and warp genres and instruments in an ever evolving stream of music. It’s been an absolute pleasure to work with Jacob Collier over the past year on this. It was very affirming to find him pushing for the same things Google Magenta Project has always valued: making tools that intrinsically value creativity and agency over just imitation, embrace the “weird corners” of AI, encouraging veterans to express themselves in new ways, while empowering novices to express themselves musically maybe for the first time. Read details at: And in the 🧵 below…

Jesse Engel

20,093 次观看 • 1 年前

Demis Hassabis thinks mathematics has a ceiling. He thinks biology is where we hit it. Hassabis: “Machine learning is the perfect description language for biology in the same way maths is for physics.” He isn’t calling AI a tool. He’s calling it a language. For four hundred years we only had one. Newton wrote gravity in it. Maxwell wrote light. Einstein wrote spacetime itself. Every law we ever pulled out of nature came back in equations, and we decided that meant nature was written in equations. It didn’t mean that. It meant the parts that surrendered first were small enough to fit the language we already spoke. Hassabis: “The expressive power of maths is not enough to understand these highly emergent dynamical systems.” Math can put a planet’s orbit on a single line. It cannot put one protein on a page. Same universe, same laws, and one fits our notation while the other refuses. Weak signals buried under noise, correlations stacked on correlations, more moving parts turning at once than any mind can hold. Biology isn’t harder than physics. It’s more expressive than the tool we brought to it. That was never a gap in our knowledge. It was a wall in our vocabulary. Hassabis is working on the far side of that wall. He calls it a virtual cell. A running simulation of a living system that no equation could ever contain. Hassabis: “Once you learn these simulators, you could maybe extract some equations from that.” He isn’t replacing math. He’s going after math we were never going to reach on our own. The model learns the system the way you learn to catch a ball, without ever solving the equation in the air. Understanding first. Formula after. That reverses the order science has followed since Newton. We started with the equation and used it to predict the system. He starts with the system and pulls the equation out of it. What if the deepest laws of life were always there, fully written, in a language we never learned to read. Math was never the language of the universe. It was the first one we spoke. Every disease we failed to cure is a sentence in that language, sitting in the open, waiting on a reader. Every person we lost to one died on the wrong side of a translation gap. The cell was never silent. We were illiterate. Hassabis isn’t building a better microscope. He’s building a second language for reality. And the first thing it’s learning to say is life.
1:37

Demis Hassabis thinks mathematics has a ceiling. He thinks biology is where we hit it. Hassabis: “Machine learning is the perfect description language for biology in the same way maths is for physics.” He isn’t calling AI a tool. He’s calling it a language. For four hundred years we only had one. Newton wrote gravity in it. Maxwell wrote light. Einstein wrote spacetime itself. Every law we ever pulled out of nature came back in equations, and we decided that meant nature was written in equations. It didn’t mean that. It meant the parts that surrendered first were small enough to fit the language we already spoke. Hassabis: “The expressive power of maths is not enough to understand these highly emergent dynamical systems.” Math can put a planet’s orbit on a single line. It cannot put one protein on a page. Same universe, same laws, and one fits our notation while the other refuses. Weak signals buried under noise, correlations stacked on correlations, more moving parts turning at once than any mind can hold. Biology isn’t harder than physics. It’s more expressive than the tool we brought to it. That was never a gap in our knowledge. It was a wall in our vocabulary. Hassabis is working on the far side of that wall. He calls it a virtual cell. A running simulation of a living system that no equation could ever contain. Hassabis: “Once you learn these simulators, you could maybe extract some equations from that.” He isn’t replacing math. He’s going after math we were never going to reach on our own. The model learns the system the way you learn to catch a ball, without ever solving the equation in the air. Understanding first. Formula after. That reverses the order science has followed since Newton. We started with the equation and used it to predict the system. He starts with the system and pulls the equation out of it. What if the deepest laws of life were always there, fully written, in a language we never learned to read. Math was never the language of the universe. It was the first one we spoke. Every disease we failed to cure is a sentence in that language, sitting in the open, waiting on a reader. Every person we lost to one died on the wrong side of a translation gap. The cell was never silent. We were illiterate. Hassabis isn’t building a better microscope. He’s building a second language for reality. And the first thing it’s learning to say is life.

Dustin

17,368 次观看 • 4 天前

Today at Meta FAIR we’re announcing three new cutting-edge developments in robotics and touch perception — and releasing a collection of artifacts to empower the community to build on this work. Details on all of this new work ➡️ 1️⃣ Meta Sparsh is the first general-purpose encoder for vision-based tactile sensing that works across many tactile sensors and many tasks. Trained on 460K+ tactile images using self-supervised learning. 2️⃣ Meta Digit 360 is a breakthrough artificial fingertip-based tactile sensor, equipped with 18+ sensing features to deliver detailed touch data with human-level precision and touch-sensing capabilities. 3️⃣ Meta Digit Plexus is a standardized platform for robotic sensor connections and interactions. It provides a hardware-software solution to integrate tactile sensors on a single robot hand and enables seamless data collection, control and analysis over a single cable. The potential impact of expanding capabilities and components like these for the open source community ranges from medical research to supply chain, manufacturing and much more. We’re excited to continue this work with the broader community.
2:35

Today at Meta FAIR we’re announcing three new cutting-edge developments in robotics and touch perception — and releasing a collection of artifacts to empower the community to build on this work. Details on all of this new work ➡️ 1️⃣ Meta Sparsh is the first general-purpose encoder for vision-based tactile sensing that works across many tactile sensors and many tasks. Trained on 460K+ tactile images using self-supervised learning. 2️⃣ Meta Digit 360 is a breakthrough artificial fingertip-based tactile sensor, equipped with 18+ sensing features to deliver detailed touch data with human-level precision and touch-sensing capabilities. 3️⃣ Meta Digit Plexus is a standardized platform for robotic sensor connections and interactions. It provides a hardware-software solution to integrate tactile sensors on a single robot hand and enables seamless data collection, control and analysis over a single cable. The potential impact of expanding capabilities and components like these for the open source community ranges from medical research to supply chain, manufacturing and much more. We’re excited to continue this work with the broader community.

AI at Meta

453,260 次观看 • 1 年前

Our fourth free DLC for Tainted Grail: The Fall of Avalon is out now! Featuring the Challenge Mode, a brand-new way to play for those who crave an extra dose of challenge and unpredictability. As you develop your character via various gameplay systems, you also have to choose harsh debuffs that make the game a ton harder. 💪💀 If you're playing on Steam, just add the DLC to your library for free. On all other platforms, it has been introduced as a part of the base game. Alongside, patch 1.22 is live, with various fixes and all-around improvements! Watch the trailer and see our recent Steam post for more details: #taintedgrail #rpg #dlc #freedlc #update
1:36

Our fourth free DLC for Tainted Grail: The Fall of Avalon is out now! Featuring the Challenge Mode, a brand-new way to play for those who crave an extra dose of challenge and unpredictability. As you develop your character via various gameplay systems, you also have to choose harsh debuffs that make the game a ton harder. 💪💀 If you're playing on Steam, just add the DLC to your library for free. On all other platforms, it has been introduced as a part of the base game. Alongside, patch 1.22 is live, with various fixes and all-around improvements! Watch the trailer and see our recent Steam post for more details: #taintedgrail #rpg #dlc #freedlc #update

Tainted Grail

34,682 次观看 • 3 个月前

We've all finished a chapter only to realize we can't remember what we just read. This points to something decades of research has confirmed: passive reading is surprisingly ineffective for learning. # The control effect Markant et al. (2014) stripped away different aspects of self-directed learning to find what actually matters. They have found that even just pressing a button when ready to see the next item—with no control over content or duration—significantly enhanced recognition memory. The researchers suggested this works because controlling when you see new information lets you coordinate stimulus presentation with your own attentional state. When you decide when to advance, your brain is better prepared to encode what comes next. # Self-pacing and strategy Tullis and Benjamin (2011) showed that self-paced learners outperformed those who studied for the same total time but couldn't control their pace. The benefit was strongest for learners who spent more time on difficult material—a "discrepancy reduction" strategy. Self-pacing isn't just about having control; it's about using it to allocate attention where it's needed. # Why agency matters DuBrow et al. (2019) found that choice are inherently rewarding. Items learned when participants could make a choice (even an inconsequential one) were remembered better, and there was a correlation between how much someone's preference increased for chosen items and how much their memory improved. Agency engages value-based brain systems that strengthen consolidation. Ding et al. (2021) extended this to incidental memory—showing that even when participants weren't trying to memorize, having control over the task improved later recognition, particularly for items processed quickly. # ChapterPal I programmed ChapterPal to implement these principles: gradual text reveal controlled by the reader, AI-generated comprehension quizzes inserted during reading, regular guess-this-blurred-term puzzles, and contextual Q&A for engaging with difficult passages. The quizzes and puzzles align with research showing that testing and guessing strengthens retention, while the self-paced reveal directly implements the minimal agency sufficient to enhance memory. # References - Tullis & Benjamin (2011). On the effectiveness of self-paced learning. Journal of Memory and Language. - Markant et al. (2014). Deconstructing the effect of self-directed study on episodic memory. Memory & Cognition. - DuBrow et al. (2019). A common mechanism underlying choice's influence on preference and memory. Psychonomic Bulletin & Review. - Ding et al. (2021). The effect of choice on intentional and incidental memory. Learning & Memory. A demo of puzzles and Q&A's on ChapterPal:
2:06

We've all finished a chapter only to realize we can't remember what we just read. This points to something decades of research has confirmed: passive reading is surprisingly ineffective for learning. # The control effect Markant et al. (2014) stripped away different aspects of self-directed learning to find what actually matters. They have found that even just pressing a button when ready to see the next item—with no control over content or duration—significantly enhanced recognition memory. The researchers suggested this works because controlling when you see new information lets you coordinate stimulus presentation with your own attentional state. When you decide when to advance, your brain is better prepared to encode what comes next. # Self-pacing and strategy Tullis and Benjamin (2011) showed that self-paced learners outperformed those who studied for the same total time but couldn't control their pace. The benefit was strongest for learners who spent more time on difficult material—a "discrepancy reduction" strategy. Self-pacing isn't just about having control; it's about using it to allocate attention where it's needed. # Why agency matters DuBrow et al. (2019) found that choice are inherently rewarding. Items learned when participants could make a choice (even an inconsequential one) were remembered better, and there was a correlation between how much someone's preference increased for chosen items and how much their memory improved. Agency engages value-based brain systems that strengthen consolidation. Ding et al. (2021) extended this to incidental memory—showing that even when participants weren't trying to memorize, having control over the task improved later recognition, particularly for items processed quickly. # ChapterPal I programmed ChapterPal to implement these principles: gradual text reveal controlled by the reader, AI-generated comprehension quizzes inserted during reading, regular guess-this-blurred-term puzzles, and contextual Q&A for engaging with difficult passages. The quizzes and puzzles align with research showing that testing and guessing strengthens retention, while the self-paced reveal directly implements the minimal agency sufficient to enhance memory. # References - Tullis & Benjamin (2011). On the effectiveness of self-paced learning. Journal of Memory and Language. - Markant et al. (2014). Deconstructing the effect of self-directed study on episodic memory. Memory & Cognition. - DuBrow et al. (2019). A common mechanism underlying choice's influence on preference and memory. Psychonomic Bulletin & Review. - Ding et al. (2021). The effect of choice on intentional and incidental memory. Learning & Memory. A demo of puzzles and Q&A's on ChapterPal:

BURKOV

39,849 次观看 • 7 个月前

I built a book. Not a PDF viewer. A real, skeuomorphic hardcover book that lives in the browser. The idea started with a simple question: What if reading a digital book actually felt like reading a real one? Since I've worked on book app designs before, I was curious to see how Claude would respond to one of those ideas that starts out sounding a little unrealistic. At first, I only wanted to recreate the experience of opening a book and flipping through its pages. Then one idea became another. I added: • Real page-flipping interactions with sound • Text highlighting in multiple colors • A voice reader • Reading reminders that continue from where you left off • Bookmarks to save important pages to share with friends later Then I thought about one of the biggest frustrations with large books. Flipping through hundreds of pages just to find page 287 isn't fun. So I added a page entry feature. Type a page number, and the book automatically flips through the pages until it gets there, just like a real book in motion. There are still more ideas I want to explore, but this was one of those projects where I kept building as the ideas came to mind. Curiosity led me here, and I like where it is taking me! Built with Claude. Designed by me. Livelink: Do let me know what you think!
1:00

I built a book. Not a PDF viewer. A real, skeuomorphic hardcover book that lives in the browser. The idea started with a simple question: What if reading a digital book actually felt like reading a real one? Since I've worked on book app designs before, I was curious to see how Claude would respond to one of those ideas that starts out sounding a little unrealistic. At first, I only wanted to recreate the experience of opening a book and flipping through its pages. Then one idea became another. I added: • Real page-flipping interactions with sound • Text highlighting in multiple colors • A voice reader • Reading reminders that continue from where you left off • Bookmarks to save important pages to share with friends later Then I thought about one of the biggest frustrations with large books. Flipping through hundreds of pages just to find page 287 isn't fun. So I added a page entry feature. Type a page number, and the book automatically flips through the pages until it gets there, just like a real book in motion. There are still more ideas I want to explore, but this was one of those projects where I kept building as the ideas came to mind. Curiosity led me here, and I like where it is taking me! Built with Claude. Designed by me. Livelink: Do let me know what you think!

Biba

28,798 次观看 • 29 天前

Mark Zuckerberg just argued that AI will force companies to hire more people. Not fewer. Three and a half billion people use Meta every day. Not one of them has a phone number to call. Mark Zuckerberg: “It’s clearly just going to automate jobs and like all these jobs are going to go away… that has not really been how the history of technology has worked.” The entire media cycle runs the same story. AI replaces workers. Industries hollow out. The human becomes unnecessary. History has never once cooperated. Voice support for 3.5 billion daily users costs between ten and twenty billion dollars a year. The math made it untouchable. So Meta never built it. AI changed the math. Zuckerberg: “Let’s say the AI can handle 90 percent of that… you’ve gotten the cost of providing that service down to one 10th.” A service that could not exist becomes standard. Overnight. The moment it goes live, the edge cases arrive. The escalations. The problems no model can close alone. Every one needs a human on the other end. Zuckerberg: “I actually think we’re probably going to go hire more customer support people.” The AI did not kill the jobs. It unlocked a service so vast the company now needs people it never would have hired. When execution costs crater, companies do not pocket the savings. They go after problems they could never afford to touch. New markets. New products. New services that were economically impossible twelve months ago. Every one creates roles that did not exist before the machine arrived. The people terrified of automation are tracking the wrong number. They count the jobs that disappear. They have no framework for the ones that haven’t been invented yet.
1:49

Mark Zuckerberg just argued that AI will force companies to hire more people. Not fewer. Three and a half billion people use Meta every day. Not one of them has a phone number to call. Mark Zuckerberg: “It’s clearly just going to automate jobs and like all these jobs are going to go away… that has not really been how the history of technology has worked.” The entire media cycle runs the same story. AI replaces workers. Industries hollow out. The human becomes unnecessary. History has never once cooperated. Voice support for 3.5 billion daily users costs between ten and twenty billion dollars a year. The math made it untouchable. So Meta never built it. AI changed the math. Zuckerberg: “Let’s say the AI can handle 90 percent of that… you’ve gotten the cost of providing that service down to one 10th.” A service that could not exist becomes standard. Overnight. The moment it goes live, the edge cases arrive. The escalations. The problems no model can close alone. Every one needs a human on the other end. Zuckerberg: “I actually think we’re probably going to go hire more customer support people.” The AI did not kill the jobs. It unlocked a service so vast the company now needs people it never would have hired. When execution costs crater, companies do not pocket the savings. They go after problems they could never afford to touch. New markets. New products. New services that were economically impossible twelve months ago. Every one creates roles that did not exist before the machine arrived. The people terrified of automation are tracking the wrong number. They count the jobs that disappear. They have no framework for the ones that haven’t been invented yet.

Dustin

369,810 次观看 • 4 个月前