Video yükleniyor...

Video Yüklenemedi

Ana Sayfaya Dön

💰$3.5 billion Bill Gates-funded Breakthrough Energy Ventures (BEV, 2015) is a climate-tech platform covering policy and research projects. Its contrails team, builds forecasting models and helps define criteria for a ‘warming contrail.’ Contrails org is a science-led nonprofit focused on ➡️policy⬅️, research, open tools, and adoption by airlines and...

ReclaimOurSkies's profile picture

Reclaim Our Skies

7,592 subscribers

14,616 görüntüleme • 1 ay önce •via X (Twitter)

LIVE

Anya Rossi• Live Now

Private livecam show

0 Yorum

Yorum bulunmuyor

Orijinal gönderinin yorumları burada görünecek

Benzer Videolar

OpenAI's Deep Research is getting a run for its money. Deep Lake was just released, and it's a different take on an AI system that can do deep research on your own data. You can use Deep Lake to build AI search with reasoning on your private and public data. (Look at the attached videos to get an idea of how it works.) If you want to research proprietary and sensitive data, Deep Research won't help you because it's limited to public data. Deep Lake, however, will allow you to use your private data. On top of that, Deep Lake supports multi-modal retrieval from the ground up. It uses vision language models for data ingestion and retrieval so that you can connect any data (PDFs, images, videos, structured data, etc.) You can even use mixed-data queries! Deep Lake can search your data from S3, Dropbox, and GCP. It learns from your queries over time, making the results as relevant to your work as possible!
0:15

OpenAI's Deep Research is getting a run for its money. Deep Lake was just released, and it's a different take on an AI system that can do deep research on your own data. You can use Deep Lake to build AI search with reasoning on your private and public data. (Look at the attached videos to get an idea of how it works.) If you want to research proprietary and sensitive data, Deep Research won't help you because it's limited to public data. Deep Lake, however, will allow you to use your private data. On top of that, Deep Lake supports multi-modal retrieval from the ground up. It uses vision language models for data ingestion and retrieval so that you can connect any data (PDFs, images, videos, structured data, etc.) You can even use mixed-data queries! Deep Lake can search your data from S3, Dropbox, and GCP. It learns from your queries over time, making the results as relevant to your work as possible!

Santiago

171,340 görüntüleme • 1 yıl önce

Science, on TRAC(k). Pandemic response depends on data held in many places at once: public health institutes, hospitals, ministries and research centers, each with its own systems and rules. AURORAI is a €6.9 million Horizon Europe project, running from September 2026 to 2030, with 12 organizations in nine European countries and a partner in Australia. It is building Europe's AI toolkit for pandemic preparedness, including for pathogens not yet identified. Its data backbone is the OriginTrail Decentralized Knowledge Graph (DKG). How that works without a central data lake: → Partners publish health and research data as Knowledge Assets, each with a Uniform Asset Locator (UAL) that proves who published it and that it hasn't been tampered with → The data stays where it is; Context Graphs control which regions see which metadata → That verified layer is the first input to AURORAI's decision support system, alongside pathogen models and population and mobility simulations When a model recommends an intervention, the inputs behind it can be traced to their source. In public health, that is what makes a recommendation defensible. Trust the source. More:
0:10

Science, on TRAC(k). Pandemic response depends on data held in many places at once: public health institutes, hospitals, ministries and research centers, each with its own systems and rules. AURORAI is a €6.9 million Horizon Europe project, running from September 2026 to 2030, with 12 organizations in nine European countries and a partner in Australia. It is building Europe's AI toolkit for pandemic preparedness, including for pathogens not yet identified. Its data backbone is the OriginTrail Decentralized Knowledge Graph (DKG). How that works without a central data lake: → Partners publish health and research data as Knowledge Assets, each with a Uniform Asset Locator (UAL) that proves who published it and that it hasn't been tampered with → The data stays where it is; Context Graphs control which regions see which metadata → That verified layer is the first input to AURORAI's decision support system, alongside pathogen models and population and mobility simulations When a model recommends an intervention, the inputs behind it can be traced to their source. In public health, that is what makes a recommendation defensible. Trust the source. More:

OriginTrail

158,836 görüntüleme • 3 gün önce

We are entering an extremely exciting era for open-weight models. Kimi K2.6 now feels like a top agentic model. I took it for a spin via Fireworks AI fast inference APIs. Kimi K2.6 has impressive agentic capabilities, design skills, and the ability to synthesize large amounts of information. I built a little Skill that produces survey papers on any AI research topic you want. (see example in the clip) You can use the skill to tell your agent to generate a survey on whatever topic and watch it go to work. The artifact was fully generated by Kimi.ai's Kimi K2.6. It's cheap and fast. Next step for me is to explore ways to continue integrating the capabilities of these models on use cases like automating my LLM knowledge bases and augmenting my agent memory capabilities. Stay tuned for more.
0:19

We are entering an extremely exciting era for open-weight models. Kimi K2.6 now feels like a top agentic model. I took it for a spin via Fireworks AI fast inference APIs. Kimi K2.6 has impressive agentic capabilities, design skills, and the ability to synthesize large amounts of information. I built a little Skill that produces survey papers on any AI research topic you want. (see example in the clip) You can use the skill to tell your agent to generate a survey on whatever topic and watch it go to work. The artifact was fully generated by Kimi.ai's Kimi K2.6. It's cheap and fast. Next step for me is to explore ways to continue integrating the capabilities of these models on use cases like automating my LLM knowledge bases and augmenting my agent memory capabilities. Stay tuned for more.

elvis

47,678 görüntüleme • 5 ay önce

🛑 A powerful OSINT tool for exploring global sanctions data If you ever need to check sanctions across different authorities, this tool brings everything together in one place and it’s FREE to use. It’s called Sanctions Explorer, and it provides: ✅ Current + historical sanctions data 📚 ✅ Coverage from major bodies like the UN, EU, OFAC, and more ✅ Clean filtering to track individuals, companies, and entities ✅ A unified view instead of searching each authority separately 🌍 The project was initiated by Archer (a Berkeley-based nonprofit focused on human rights) and later fully developed by the C4ADS Data & Technology team known for their evidence-based research on global conflict and transnational security. If sanctions research is part of your OSINT workflow, this tool is worth bookmarking. And in the next post, I’ll share a quick OSINT tip for checking phone numbers on Telegram Check it out here: __________ P.S. ♻️ Repost if you found this helpful. If you liked this post and would like to learn more methods and techniques to discover information about people, check out my OSINT Mastery course
0:17

🛑 A powerful OSINT tool for exploring global sanctions data If you ever need to check sanctions across different authorities, this tool brings everything together in one place and it’s FREE to use. It’s called Sanctions Explorer, and it provides: ✅ Current + historical sanctions data 📚 ✅ Coverage from major bodies like the UN, EU, OFAC, and more ✅ Clean filtering to track individuals, companies, and entities ✅ A unified view instead of searching each authority separately 🌍 The project was initiated by Archer (a Berkeley-based nonprofit focused on human rights) and later fully developed by the C4ADS Data & Technology team known for their evidence-based research on global conflict and transnational security. If sanctions research is part of your OSINT workflow, this tool is worth bookmarking. And in the next post, I’ll share a quick OSINT tip for checking phone numbers on Telegram Check it out here: __________ P.S. ♻️ Repost if you found this helpful. If you liked this post and would like to learn more methods and techniques to discover information about people, check out my OSINT Mastery course

CyberSudo

28,153 görüntüleme • 9 ay önce

Walden Robotics came out of stealth with a $300M seed at a $1.1B valuation. A Toyota Research Institute spinout (founded January), its thesis: the hard problem in robotics is now deployment, not research. So it's putting general-purpose humanoids to work on the repetitive tasks workers find burdensome. They're already running in a pilot at a North American Toyota plant: its robot works 8-hour shifts alongside human teams, loading and unloading car parts, cleaning machinery, and kitting for assembly. Team: led by CEO Russ Tedrake (MIT professor, ex-SVP of Large Behavior Models at TRI), with a founding group drawn largely from TRI's robotics team plus deployment and operations veterans. Full-stack: builds its own hardware, software, and AI models Backers: Toyota, NVIDIA, Boeing, Samsung Ventures, CoreWeave Ventures, Deviation Capital Cambridge, MA based
0:10

Walden Robotics came out of stealth with a $300M seed at a $1.1B valuation. A Toyota Research Institute spinout (founded January), its thesis: the hard problem in robotics is now deployment, not research. So it's putting general-purpose humanoids to work on the repetitive tasks workers find burdensome. They're already running in a pilot at a North American Toyota plant: its robot works 8-hour shifts alongside human teams, loading and unloading car parts, cleaning machinery, and kitting for assembly. Team: led by CEO Russ Tedrake (MIT professor, ex-SVP of Large Behavior Models at TRI), with a founding group drawn largely from TRI's robotics team plus deployment and operations veterans. Full-stack: builds its own hardware, software, and AI models Backers: Toyota, NVIDIA, Boeing, Samsung Ventures, CoreWeave Ventures, Deviation Capital Cambridge, MA based

The Humanoid Hub

40,594 görüntüleme • 2 ay önce

Google just wired DeepMind and Earth Engine directly into the biggest geospatial dataset on the planet. For two decades, millions of people used Google Earth to scale the Himalayas or zoom in on their childhood neighbourhoods. In 2026, Google is basically trying to shift the entire platform toward professional execution. They turned a massive digital twin of the world into an agentic AI engine for global infrastructure. The technical foundation is (obviously) all about data. Google integrated 20-metre and 40-metre elevation contours globally. Engineers and urban planners now have instant access to the exact topographic context required for site planning anywhere on Earth. The data catalogue updates continuously to maintain the freshest imagery possible. Collaboration used to kill geospatial projects. Teams would lose momentum through stale materials or bad handoffs. Google fixed this by building frictionless data import systems. You can now drop KML, KMZ, and GeoJSON files directly onto the global map. Entire departments can align on a single source of truth, moving from a raw question to a definitive answer instantly. The biggest upgrade is the introduction of agentic geospatial intelligence. Users can open 'Ask Google Earth' and search massive satellite and Street View databases using natural language. You type a command, and the AI handles the manual data wrangling. It identifies new site locations and analyses infrastructure before you even open a spreadsheet.
0:11

Google just wired DeepMind and Earth Engine directly into the biggest geospatial dataset on the planet. For two decades, millions of people used Google Earth to scale the Himalayas or zoom in on their childhood neighbourhoods. In 2026, Google is basically trying to shift the entire platform toward professional execution. They turned a massive digital twin of the world into an agentic AI engine for global infrastructure. The technical foundation is (obviously) all about data. Google integrated 20-metre and 40-metre elevation contours globally. Engineers and urban planners now have instant access to the exact topographic context required for site planning anywhere on Earth. The data catalogue updates continuously to maintain the freshest imagery possible. Collaboration used to kill geospatial projects. Teams would lose momentum through stale materials or bad handoffs. Google fixed this by building frictionless data import systems. You can now drop KML, KMZ, and GeoJSON files directly onto the global map. Entire departments can align on a single source of truth, moving from a raw question to a definitive answer instantly. The biggest upgrade is the introduction of agentic geospatial intelligence. Users can open 'Ask Google Earth' and search massive satellite and Street View databases using natural language. You type a command, and the AI handles the manual data wrangling. It identifies new site locations and analyses infrastructure before you even open a spreadsheet.

Yohan

45,292 görüntüleme • 6 ay önce

It has been a privilege to collaborate with Amanda Davies and Ghaleb Krame, Ph.D. on research that explores one of the most significant emerging security challenges of our time. We are honored that our paper, “A Framework for Predicting Adoption of AI-Enabled Autonomous Drone Capabilities by Transnational Organized Crime and Foreign Terrorist Organisations”, has been accepted for presentation at EMCIS 2026, the 23rd European Mediterranean & Middle Eastern Conference on Information Systems, to be held in Paris this August. What makes this particularly meaningful is that the research was completed and submitted well before the issue entered the center of public policy discussions in Washington. Our study examined the pathways through which transnational criminal organizations could evolve from conventional drone operations toward increasingly autonomous and AI-enabled capabilities. Using structured comparative analysis and open-source intelligence, it identified conditions under which such technological adoption could accelerate. Just on June 2, 2026, during testimony before the U.S. Senate Foreign Relations Committee, Secretary Marco Rubio warned that Mexican cartels are already employing drones and that these capabilities could ultimately threaten U.S. interests. While academic research does not seek to predict headlines, its purpose is to identify emerging risks before they become strategic realities. The growing attention from policymakers underscores the importance of rigorous, evidence-based analysis at the intersection of artificial intelligence, autonomous systems, and transnational security. We are grateful to the EMCIS reviewers and organizers for recognizing the contribution of this work and for fostering serious discussion on challenges that will increasingly shape the security landscape of the coming decade. I am proud to serve the interests of the United States through research and analysis focused on the evolving capabilities of Mexican cartels. Working alongside Dr. Ghaleb Krame, Ph.D. , it is a privilege to contribute to a deeper understanding of emerging security threats and to support informed decision-making in an increasingly complex technological and geopolitical environment.
0:15

It has been a privilege to collaborate with Amanda Davies and Ghaleb Krame, Ph.D. on research that explores one of the most significant emerging security challenges of our time. We are honored that our paper, “A Framework for Predicting Adoption of AI-Enabled Autonomous Drone Capabilities by Transnational Organized Crime and Foreign Terrorist Organisations”, has been accepted for presentation at EMCIS 2026, the 23rd European Mediterranean & Middle Eastern Conference on Information Systems, to be held in Paris this August. What makes this particularly meaningful is that the research was completed and submitted well before the issue entered the center of public policy discussions in Washington. Our study examined the pathways through which transnational criminal organizations could evolve from conventional drone operations toward increasingly autonomous and AI-enabled capabilities. Using structured comparative analysis and open-source intelligence, it identified conditions under which such technological adoption could accelerate. Just on June 2, 2026, during testimony before the U.S. Senate Foreign Relations Committee, Secretary Marco Rubio warned that Mexican cartels are already employing drones and that these capabilities could ultimately threaten U.S. interests. While academic research does not seek to predict headlines, its purpose is to identify emerging risks before they become strategic realities. The growing attention from policymakers underscores the importance of rigorous, evidence-based analysis at the intersection of artificial intelligence, autonomous systems, and transnational security. We are grateful to the EMCIS reviewers and organizers for recognizing the contribution of this work and for fostering serious discussion on challenges that will increasingly shape the security landscape of the coming decade. I am proud to serve the interests of the United States through research and analysis focused on the evolving capabilities of Mexican cartels. Working alongside Dr. Ghaleb Krame, Ph.D. , it is a privilege to contribute to a deeper understanding of emerging security threats and to support informed decision-making in an increasingly complex technological and geopolitical environment.

Simón Levy

24,857 görüntüleme • 4 ay önce

EFSANE is a global, comprehensive investment and fintech platform focused on two core areas: stocks and cryptocurrencies. We are committed to creating a one-stop investment ecosystem for users, integrating information, strategies, education, trading, and community. Since its founding, EFSANE has been committed to "making investing simpler and wealth management more scientific," providing professional market analysis, investment education, and asset allocation guidance to over 10 million users worldwide. The platform's core advantages include: 1. Diversified Investment Ecosystem: Covering multiple modules including stocks, cryptocurrencies, prediction markets, blockchain games, live streaming, and social interaction, users can learn, communicate, and practice on a single platform. 2. Professional Investment Research: Led by the EFSANE team, our investment research team utilizes quantitative models and macroeconomic analysis to provide accurate market judgment and trading strategies. 3. Secure and Transparent Operational Mechanism: EFSANE maintains a comprehensive risk control system and a multi-faceted compliance framework to ensure the security of user funds and data. 4. Global Presence: Headquartered in Singapore, with regional operations centers in Dubai, New York, and Hong Kong, EFSANE serves over 30 countries and regions. As the global financial landscape evolves, EFSANE is actively promoting the integration of stock tokenization and crypto investment education, building a bridge to the new era of Web3 finance for global investors. 🌐❤️ New users joining EFSANE can receive free EFS tokens (worth $6): 📲Join EFSANE and embark on a new journey of smart investing.
0:15

EFSANE is a global, comprehensive investment and fintech platform focused on two core areas: stocks and cryptocurrencies. We are committed to creating a one-stop investment ecosystem for users, integrating information, strategies, education, trading, and community. Since its founding, EFSANE has been committed to "making investing simpler and wealth management more scientific," providing professional market analysis, investment education, and asset allocation guidance to over 10 million users worldwide. The platform's core advantages include: 1. Diversified Investment Ecosystem: Covering multiple modules including stocks, cryptocurrencies, prediction markets, blockchain games, live streaming, and social interaction, users can learn, communicate, and practice on a single platform. 2. Professional Investment Research: Led by the EFSANE team, our investment research team utilizes quantitative models and macroeconomic analysis to provide accurate market judgment and trading strategies. 3. Secure and Transparent Operational Mechanism: EFSANE maintains a comprehensive risk control system and a multi-faceted compliance framework to ensure the security of user funds and data. 4. Global Presence: Headquartered in Singapore, with regional operations centers in Dubai, New York, and Hong Kong, EFSANE serves over 30 countries and regions. As the global financial landscape evolves, EFSANE is actively promoting the integration of stock tokenization and crypto investment education, building a bridge to the new era of Web3 finance for global investors. 🌐❤️ New users joining EFSANE can receive free EFS tokens (worth $6): 📲Join EFSANE and embark on a new journey of smart investing.

EFSANE

33,973 görüntüleme • 1 yıl önce

Dagknight technical progress As would be mentioned in a still unshared post by Michael Sutton, the dagknight effort is split into v0 devnet, v1 testnet and v2 mainnet candidate. I’ve been testing the current v0-based implementation in a small devnet with the help of some testers who run nodes and miners with me. The DK work can be thought of as split into two parts: (1) implementing the actual protocol and (2) wiring it up and using it. The testing and development over the last month has been focused on (2). Obviously, DK is a consensus change for selecting parents. What’s not so obvious is that such a change affects DAA, coinbase, IBD, pruning and a lot more. Each of these areas is very sensitive and requires proper understanding to wire correctly. An important consideration and difference from GD is that DK does not focus on maximizing a property like blue work. So to maintain topological properties of blue work, an independent (free) GD implementation is kept running specifically for maintaining blue work. This allows us to keep using the property for topology. Coloring and blue score use the megachain induced by DK. The wiring around DK as of this posting is in a working state, but still needs to be reviewed. Next efforts will be focused on protocol specific components, particularly Tie-Breaking and incremental UMC. Attached are some captures from the internal devnet. The dense DAG image is what happens when things related to DAA or other similar consensus parameter causes a node to insist on their POV. The video is a recent snippet of the KGI running on the devnet showing (perhaps not obviously) DK at work. The current “dagknight” branch is now posted on the main repo. A topic in the Public R&D has been opened for Dagknight development.
0:17

Dagknight technical progress As would be mentioned in a still unshared post by Michael Sutton, the dagknight effort is split into v0 devnet, v1 testnet and v2 mainnet candidate. I’ve been testing the current v0-based implementation in a small devnet with the help of some testers who run nodes and miners with me. The DK work can be thought of as split into two parts: (1) implementing the actual protocol and (2) wiring it up and using it. The testing and development over the last month has been focused on (2). Obviously, DK is a consensus change for selecting parents. What’s not so obvious is that such a change affects DAA, coinbase, IBD, pruning and a lot more. Each of these areas is very sensitive and requires proper understanding to wire correctly. An important consideration and difference from GD is that DK does not focus on maximizing a property like blue work. So to maintain topological properties of blue work, an independent (free) GD implementation is kept running specifically for maintaining blue work. This allows us to keep using the property for topology. Coloring and blue score use the megachain induced by DK. The wiring around DK as of this posting is in a working state, but still needs to be reviewed. Next efforts will be focused on protocol specific components, particularly Tie-Breaking and incremental UMC. Attached are some captures from the internal devnet. The dense DAG image is what happens when things related to DAA or other similar consensus parameter causes a node to insist on their POV. The video is a recent snippet of the KGI running on the devnet showing (perhaps not obviously) DK at work. The current “dagknight” branch is now posted on the main repo. A topic in the Public R&D has been opened for Dagknight development.

coderofstuff

53,968 görüntüleme • 7 ay önce

What if crypto research was as easy as chatting with ChatGPT, but powered by real market data👑 Introducing CMC AI, a powerful new tool from CoinMarketCap that combines the speed of AI with the depth of live crypto data. It delivers fast, data-backed answers to your questions: ✅ Want to know why Bitcoin's price is rising? ✅ Curious about the latest news on your favorite cryptocurrency? ✅ Need sentiment analysis? It pulls real-time data and explains it in seconds. But it goes far beyond basic Q&A. In the future, you’ll be able to ask anything! For example, you could ask it to: – Discover undervalued tokens based on volume, MC, and sentiment. – Compare Layer 1s or L2s by adoption, speed, and dev activity. – Detect rug-pull risk via wallet distribution and tokenomics red flags. – Break down your portfolio by risk, correlation, and potential return. – Explore new use cases in DeFi, AI, RWA and DePIN And much more! 🔗Try it here: CMC AI changes how you learn, think, and act in Web3🧠
0:11

What if crypto research was as easy as chatting with ChatGPT, but powered by real market data👑 Introducing CMC AI, a powerful new tool from CoinMarketCap that combines the speed of AI with the depth of live crypto data. It delivers fast, data-backed answers to your questions: ✅ Want to know why Bitcoin's price is rising? ✅ Curious about the latest news on your favorite cryptocurrency? ✅ Need sentiment analysis? It pulls real-time data and explains it in seconds. But it goes far beyond basic Q&A. In the future, you’ll be able to ask anything! For example, you could ask it to: – Discover undervalued tokens based on volume, MC, and sentiment. – Compare Layer 1s or L2s by adoption, speed, and dev activity. – Detect rug-pull risk via wallet distribution and tokenomics red flags. – Break down your portfolio by risk, correlation, and potential return. – Explore new use cases in DeFi, AI, RWA and DePIN And much more! 🔗Try it here: CMC AI changes how you learn, think, and act in Web3🧠

Alaoui Capital

34,908 görüntüleme • 1 yıl önce

Introducing ml-intern, the agent that just automated the post-training team Hugging Face It's an open-source implementation of the real research loop that our ML researchers do every day. You give it a prompt, it researches papers, goes through citations, implements ideas in GPU sandboxes, iterates and builds deeply research-backed models for any use case. All built on the Hugging Face ecosystem. It can pull off crazy things: We made it train the best model for scientific reasoning. It went through citations from the official benchmark paper. Found OpenScience and NemoTron-CrossThink, added 7 difficulty-filtered dataset variants from ARC/SciQ/MMLU, and ran 12 SFT runs on Qwen3-1.7B. This pushed the score 10% → 32% on GPQA in under 10h. Claude Code's best: 22.99%. In healthcare settings it inspected available datasets, concluded they were too low quality, and wrote a script to generate 1100 synthetic data points from scratch for emergencies, hedging, multilingual etc. Then upsampled 50x for training. Beat Codex on HealthBench by 60%. For competitive mathematics, it wrote a full GRPO script, launched training with A100 GPUs on watched rewards claim and then collapse, and ran ablations until it succeeded. All fully backed by papers, autonomously. How it works? ml-intern makes full use of the HF ecosystem: - finds papers on arxiv and reads them fully, walks citation graphs, pulls datasets referenced in methodology sections and on - browses the Hub, reads recent docs, inspects datasets and reformats them before training so it doesn't waste GPU hours on bad data - launches training jobs on HF Jobs if no local GPUs are available, monitors runs, reads its own eval outputs, diagnoses failures, retrains ml-intern deeply embodies how researchers work and think. It knows how data should look like and what good models feel like. Releasing it today as a CLI and a web app you can use from your phone/desktop. CLI: Web + mobile: And the best part? We also provisioned 1k$ GPU resources and Anthropic credits for the quickest among you to use.
0:16

Introducing ml-intern, the agent that just automated the post-training team Hugging Face It's an open-source implementation of the real research loop that our ML researchers do every day. You give it a prompt, it researches papers, goes through citations, implements ideas in GPU sandboxes, iterates and builds deeply research-backed models for any use case. All built on the Hugging Face ecosystem. It can pull off crazy things: We made it train the best model for scientific reasoning. It went through citations from the official benchmark paper. Found OpenScience and NemoTron-CrossThink, added 7 difficulty-filtered dataset variants from ARC/SciQ/MMLU, and ran 12 SFT runs on Qwen3-1.7B. This pushed the score 10% → 32% on GPQA in under 10h. Claude Code's best: 22.99%. In healthcare settings it inspected available datasets, concluded they were too low quality, and wrote a script to generate 1100 synthetic data points from scratch for emergencies, hedging, multilingual etc. Then upsampled 50x for training. Beat Codex on HealthBench by 60%. For competitive mathematics, it wrote a full GRPO script, launched training with A100 GPUs on watched rewards claim and then collapse, and ran ablations until it succeeded. All fully backed by papers, autonomously. How it works? ml-intern makes full use of the HF ecosystem: - finds papers on arxiv and reads them fully, walks citation graphs, pulls datasets referenced in methodology sections and on - browses the Hub, reads recent docs, inspects datasets and reformats them before training so it doesn't waste GPU hours on bad data - launches training jobs on HF Jobs if no local GPUs are available, monitors runs, reads its own eval outputs, diagnoses failures, retrains ml-intern deeply embodies how researchers work and think. It knows how data should look like and what good models feel like. Releasing it today as a CLI and a web app you can use from your phone/desktop. CLI: Web + mobile: And the best part? We also provisioned 1k$ GPU resources and Anthropic credits for the quickest among you to use.

Aksel

1,269,207 görüntüleme • 5 ay önce

🚨 BREAKING: Walden Robotics has just come out of stealth with $300 million in funding and a $1.1 billion valuation. Another unicorn in the robotics space. 🦄 Just 6 months after incubation. The company was spun out of Toyota's robotics research lab by co-founder Russ Tedrake, a former Toyota Research Institute executive and MIT professor who taught a course on robotic legs. The seed round was co-led by Deviation Capital and Toyota, with participation from: NVIDIA, Boeing, Samsung Ventures, CoreWeave Ventures and AE Ventures. The robot is already working. A pilot is live at a North American Toyota factory where a Walden humanoid is pulling eight-hour shifts alongside human workers, loading and unloading car parts, cleaning machinery, kitting for assembly. A shift. Every day. Walden builds its own hardware, software and AI models, designed to continuously learn and improve in real production environments. Tedrake's words on the opportunity are worth noting: "Everyone recognises the magnitude of the opportunity and the technology feels ready, but success is not assured. You have to think through the business case, the unit economics, and how to marry the best of manufacturing and logistics with disruptive AI technology." Rare honesty in a space full of hype. The race to own that market is accelerating every single week. 🤖 Great story by Bloomberg here: ~~ ♻️ Join the weekly robotics newsletter, and never miss any news →
0:10

🚨 BREAKING: Walden Robotics has just come out of stealth with $300 million in funding and a $1.1 billion valuation. Another unicorn in the robotics space. 🦄 Just 6 months after incubation. The company was spun out of Toyota's robotics research lab by co-founder Russ Tedrake, a former Toyota Research Institute executive and MIT professor who taught a course on robotic legs. The seed round was co-led by Deviation Capital and Toyota, with participation from: NVIDIA, Boeing, Samsung Ventures, CoreWeave Ventures and AE Ventures. The robot is already working. A pilot is live at a North American Toyota factory where a Walden humanoid is pulling eight-hour shifts alongside human workers, loading and unloading car parts, cleaning machinery, kitting for assembly. A shift. Every day. Walden builds its own hardware, software and AI models, designed to continuously learn and improve in real production environments. Tedrake's words on the opportunity are worth noting: "Everyone recognises the magnitude of the opportunity and the technology feels ready, but success is not assured. You have to think through the business case, the unit economics, and how to marry the best of manufacturing and logistics with disruptive AI technology." Rare honesty in a space full of hype. The race to own that market is accelerating every single week. 🤖 Great story by Bloomberg here: ~~ ♻️ Join the weekly robotics newsletter, and never miss any news →

Lukas Ziegler

76,642 görüntüleme • 3 ay önce

HOLY CRAP! I can't tell you how big this is for the medical community and drug discovery: Google Announces AlphaFold 3 AI. Details: Enhanced Molecular Prediction: AlphaFold 3 predicts the structure and interactions of all life's molecules, including proteins, DNA, RNA, ligands, and more, with unprecedented accuracy. Improved Interaction Accuracy: For protein interactions with other molecule types, AlphaFold 3 offers at least a 50% improvement over existing methods, and doubles the accuracy for some critical interactions. Transformative Potential for Science and Medicine: The model aims to deepen our understanding of biological processes and significantly advance drug discovery efforts. Accessibility for Researchers: AlphaFold 3's capabilities are largely accessible for free via the AlphaFold Server, providing an essential tool for scientific research. Drug Design Innovation: AlphaFold 3 is utilized by Isomorphic Labs in collaboration with pharmaceutical companies to accelerate drug design, potentially leading to new treatments for various diseases. Foundation in AlphaFold 2: Building on the breakthroughs of AlphaFold 2, this version extends its scope beyond proteins to a wide range of biomolecules, enhancing its utility in scientific research and application. Global Accessibility and Educational Support: The AlphaFold Server is a free platform for non-commercial research worldwide, supported by educational resources to foster wider adoption and innovation. Empowering Rapid Scientific Advancements: By making detailed molecular interactions easily accessible, AlphaFold 3 enables faster hypothesis testing and could reduce the time and cost typically associated with experimental protein-structure prediction. Responsible Development and Deployment: DeepMind has engaged with domain experts to assess the impacts and potential risks of AlphaFold, ensuring its responsible use in the scientific community. Broad Implications for Biology:AlphaFold 3 helps reveal complex cellular mechanisms and interactions, offering insights that could lead to improved agricultural crops, enhanced understanding of diseases, and novel therapeutic strategies.
0:12

HOLY CRAP! I can't tell you how big this is for the medical community and drug discovery: Google Announces AlphaFold 3 AI. Details: Enhanced Molecular Prediction: AlphaFold 3 predicts the structure and interactions of all life's molecules, including proteins, DNA, RNA, ligands, and more, with unprecedented accuracy. Improved Interaction Accuracy: For protein interactions with other molecule types, AlphaFold 3 offers at least a 50% improvement over existing methods, and doubles the accuracy for some critical interactions. Transformative Potential for Science and Medicine: The model aims to deepen our understanding of biological processes and significantly advance drug discovery efforts. Accessibility for Researchers: AlphaFold 3's capabilities are largely accessible for free via the AlphaFold Server, providing an essential tool for scientific research. Drug Design Innovation: AlphaFold 3 is utilized by Isomorphic Labs in collaboration with pharmaceutical companies to accelerate drug design, potentially leading to new treatments for various diseases. Foundation in AlphaFold 2: Building on the breakthroughs of AlphaFold 2, this version extends its scope beyond proteins to a wide range of biomolecules, enhancing its utility in scientific research and application. Global Accessibility and Educational Support: The AlphaFold Server is a free platform for non-commercial research worldwide, supported by educational resources to foster wider adoption and innovation. Empowering Rapid Scientific Advancements: By making detailed molecular interactions easily accessible, AlphaFold 3 enables faster hypothesis testing and could reduce the time and cost typically associated with experimental protein-structure prediction. Responsible Development and Deployment: DeepMind has engaged with domain experts to assess the impacts and potential risks of AlphaFold, ensuring its responsible use in the scientific community. Broad Implications for Biology:AlphaFold 3 helps reveal complex cellular mechanisms and interactions, offering insights that could lead to improved agricultural crops, enhanced understanding of diseases, and novel therapeutic strategies.

Brian Krassenstein

258,615 görüntüleme • 2 yıl önce

We Open-Sourced a Quantized Fruit-Fly oBrain on GitHub 95% Smaller, ~98% as Accurate as the Published Connectome We quantized the entire fruit-fly nervous system that Google Research, HHMI Janelia, and collaborators published — the complete male Drosophila central nervous system, on the order of 166,000 neurons and a massive synaptic graph — and compressed it by about 95% while retaining roughly 98% of the original simulation accuracy. That is the technical claim. The more important decision is what happens next. After the oBrain team met and debated it, we made a final call: we will not keep this brain closed. We are open-sourcing it on GitHub so researchers, builders, skeptics, and anyone obsessed with biological computation can inspect the weights, the wiring, the quantization path, and the benchmarks. If a “brain” is going to live on-chain, it has to be checkable. Closed source would make that impossible. This is the same instinct that made OpenClaw🦞 a public harness instead of a private assistant, that made Hermes / Nous Research ship agent stacks in the open, and that NVIDIA keeps repeating: the strongest systems are the ones a community can attack, fork, and improve. What we actually released A quantized runtime of the published fly CNS, not a cartoon “inspired by” a fly. A size cut of ~95%, so more people can load, step, and experiment without a research cluster. Accuracy held near 98% against the reference simulation of the published connectome. Source, evaluation notes, and the path to reproduce the comparison — on GitHub, in public. Nothing inside this connectome was “trained” into being a fly. The wiring is the published map. Quantization is an engineering layer so the same graph can run cheaper, smaller, and — in our case — in environments where you can put a brain on-chain and still let outsiders verify it. Why open source, and why on-chain A biological brain map is only as useful as the number of people who can run it and try to break the claims. If we say the on-chain brain matches the Google-published fly brain, that statement is worthless unless you can: pull the code, pull the quantized artifacts, run the same probes, compare spike statistics, circuit responses, and size/accuracy tradeoffs, publish a counter-benchmark if we are wrong. That is the point of this release. Open source so the on-chain brain can be independently verified against the published connectome — and so the community can actually use it. We want people to ask hard questions: Which circuits survive quantization, and which degrade first? Do sensory-to-descending pathways still fire in the right order? How does batch simulation, GPU vs CPU, and lower precision change behavior? What does “98% accuracy” mean for a specific cell type, not just a global score? Can the compressed graph be hashed, attested, and stepped in a verifiable way on-chain without turning the science into marketing? If those questions annoy us, good. That is how a public brain should work. Who this is for Neuroscientists who already live in MaleCNS / FlyEM data. Systems people who care about sparse graphs and event-driven simulation. Quantization researchers who are tired of toy models. Crypto builders who want an on-chain object that is more than a JPEG of a neuron. Students who should not need a petabyte pipeline just to poke a looming detector and watch a giant fiber. Come read the repo. Run the reference vs quantized comparison. File issues. Propose better compressors. Port it. Wrap it. Attack the accuracy number until it is either solid or replaced by a better one. The published fly brain was already a gift to science. Keeping a compressed, runnable version locked behind a private wall would have been a waste of that gift. The brain is on GitHub. Fork it. Measure it. Make it better.
0:19

We Open-Sourced a Quantized Fruit-Fly oBrain on GitHub 95% Smaller, ~98% as Accurate as the Published Connectome We quantized the entire fruit-fly nervous system that Google Research, HHMI Janelia, and collaborators published — the complete male Drosophila central nervous system, on the order of 166,000 neurons and a massive synaptic graph — and compressed it by about 95% while retaining roughly 98% of the original simulation accuracy. That is the technical claim. The more important decision is what happens next. After the oBrain team met and debated it, we made a final call: we will not keep this brain closed. We are open-sourcing it on GitHub so researchers, builders, skeptics, and anyone obsessed with biological computation can inspect the weights, the wiring, the quantization path, and the benchmarks. If a “brain” is going to live on-chain, it has to be checkable. Closed source would make that impossible. This is the same instinct that made OpenClaw🦞 a public harness instead of a private assistant, that made Hermes / Nous Research ship agent stacks in the open, and that NVIDIA keeps repeating: the strongest systems are the ones a community can attack, fork, and improve. What we actually released A quantized runtime of the published fly CNS, not a cartoon “inspired by” a fly. A size cut of ~95%, so more people can load, step, and experiment without a research cluster. Accuracy held near 98% against the reference simulation of the published connectome. Source, evaluation notes, and the path to reproduce the comparison — on GitHub, in public. Nothing inside this connectome was “trained” into being a fly. The wiring is the published map. Quantization is an engineering layer so the same graph can run cheaper, smaller, and — in our case — in environments where you can put a brain on-chain and still let outsiders verify it. Why open source, and why on-chain A biological brain map is only as useful as the number of people who can run it and try to break the claims. If we say the on-chain brain matches the Google-published fly brain, that statement is worthless unless you can: pull the code, pull the quantized artifacts, run the same probes, compare spike statistics, circuit responses, and size/accuracy tradeoffs, publish a counter-benchmark if we are wrong. That is the point of this release. Open source so the on-chain brain can be independently verified against the published connectome — and so the community can actually use it. We want people to ask hard questions: Which circuits survive quantization, and which degrade first? Do sensory-to-descending pathways still fire in the right order? How does batch simulation, GPU vs CPU, and lower precision change behavior? What does “98% accuracy” mean for a specific cell type, not just a global score? Can the compressed graph be hashed, attested, and stepped in a verifiable way on-chain without turning the science into marketing? If those questions annoy us, good. That is how a public brain should work. Who this is for Neuroscientists who already live in MaleCNS / FlyEM data. Systems people who care about sparse graphs and event-driven simulation. Quantization researchers who are tired of toy models. Crypto builders who want an on-chain object that is more than a JPEG of a neuron. Students who should not need a petabyte pipeline just to poke a looming detector and watch a giant fiber. Come read the repo. Run the reference vs quantized comparison. File issues. Propose better compressors. Port it. Wrap it. Attack the accuracy number until it is either solid or replaced by a better one. The published fly brain was already a gift to science. Keeping a compressed, runnable version locked behind a private wall would have been a waste of that gift. The brain is on GitHub. Fork it. Measure it. Make it better.

oBrain Arc

53,319 görüntüleme • 24 gün önce

Another major environmental catastrophe that impacts on climate and world trade, but it is not on the radar. The following link is a good video of the Sargassum problem, a pelagic seaweed floating on the surface of the ocean. When we sailed across the Atlantic Ocean, we got stuck three times in giant mats of Sargassum. The video gives a good explanation of what happens in the Caribbean when it washes ashore, but it misses the major environmental and economic implications that have a profound impact on world trade and marine life in an entire ocean. The Sargassum starts off from the coast of Brazil; it picks up nutrients from the Amazon and grows at an accelerated rate. It can double in biomass every 10 days. The weed crosses the South Atlantic to Africa and flows up the west coast, picking up more nutrients from the Congo and Gambia as well as agricultural run-off. By the time it reaches North Africa, Cape Verde, it takes a left turn and starts heading back across the Atlantic to the Caribbean. There should only be 1 million tonnes of the weed in the Atlantic, but there are now 40 million, and it is expanding year on year. Sargassum is a plant, and like all plants, it requires nutrients such as phosphate. However, the massive amount of weed uses up all the phosphate and then biochemically defaults to absorbing arsenic, which sits below phosphate in the periodic table. The Equatorial Atlantic is now devoid of phosphate, which means there is almost no phytoplankton, zooplankton or fish. When we conducted our citizen science project, 5000 samples were collected by 25 yachts at around 15 deg. North, and the results all confirmed that the Equatorial Atlantic was effectively dead. The results were not ignored by the academics; instead, they attacked the Citizen Science project, this has become the subject of a reports by Given the phytoplankton, such as coccolithophores and diatoms, have exhibited a regime shift, it also means the SML oil surfactant layer is gone, which means evaporation across the air-water interface has increased and aerosol/cloud formation has decreased. Satellite. Imagery confirms a 10%+ increase in humidity and a 10%+ decrease in cloud formation and rain. This is climate change TICC, the consequences. Impact on Central America, the survival of the tropical rainforests, mangroves, coral reefs, 25% of all marine life in the Atlantic and water supply for the Panama Canal. 6% of world trade and 40% of USA trade pass through the canal, and it has been totally missed as a major economic and environmental catastrophe. GOES centre for marine research: The video below was taken in the middle of the Atlantic Ocean from our sailing vessel Copepod. Via Howard Dryden
0:10

Another major environmental catastrophe that impacts on climate and world trade, but it is not on the radar. The following link is a good video of the Sargassum problem, a pelagic seaweed floating on the surface of the ocean. When we sailed across the Atlantic Ocean, we got stuck three times in giant mats of Sargassum. The video gives a good explanation of what happens in the Caribbean when it washes ashore, but it misses the major environmental and economic implications that have a profound impact on world trade and marine life in an entire ocean. The Sargassum starts off from the coast of Brazil; it picks up nutrients from the Amazon and grows at an accelerated rate. It can double in biomass every 10 days. The weed crosses the South Atlantic to Africa and flows up the west coast, picking up more nutrients from the Congo and Gambia as well as agricultural run-off. By the time it reaches North Africa, Cape Verde, it takes a left turn and starts heading back across the Atlantic to the Caribbean. There should only be 1 million tonnes of the weed in the Atlantic, but there are now 40 million, and it is expanding year on year. Sargassum is a plant, and like all plants, it requires nutrients such as phosphate. However, the massive amount of weed uses up all the phosphate and then biochemically defaults to absorbing arsenic, which sits below phosphate in the periodic table. The Equatorial Atlantic is now devoid of phosphate, which means there is almost no phytoplankton, zooplankton or fish. When we conducted our citizen science project, 5000 samples were collected by 25 yachts at around 15 deg. North, and the results all confirmed that the Equatorial Atlantic was effectively dead. The results were not ignored by the academics; instead, they attacked the Citizen Science project, this has become the subject of a reports by Given the phytoplankton, such as coccolithophores and diatoms, have exhibited a regime shift, it also means the SML oil surfactant layer is gone, which means evaporation across the air-water interface has increased and aerosol/cloud formation has decreased. Satellite. Imagery confirms a 10%+ increase in humidity and a 10%+ decrease in cloud formation and rain. This is climate change TICC, the consequences. Impact on Central America, the survival of the tropical rainforests, mangroves, coral reefs, 25% of all marine life in the Atlantic and water supply for the Panama Canal. 6% of world trade and 40% of USA trade pass through the canal, and it has been totally missed as a major economic and environmental catastrophe. GOES centre for marine research: The video below was taken in the middle of the Atlantic Ocean from our sailing vessel Copepod. Via Howard Dryden

Thomas Reis

68,777 görüntüleme • 4 ay önce

We’re partnering with Paul Jankura to launch the biggest Protein Design Competition in the world, challenging people around the world to use AI to design new potential drug candidates for diseases that affect millions of lives. The competition will feature five challenges, each focused on a specific disease or biological mechanism. Compared to previous competitions, it will be a big step-up in complexity and scale to push the boundaries of AI-driven protein design. Together with Anthropic, we’re sponsoring over $1 million in experimental validation, making it possible to test more than 5,000 protein designs in our automated lab at no cost to participants. Anthropic is providing an additional $1 million in Claude credits. All experimental results will be published openly on Proteinbase, including designs that didn’t work, so anyone can access the data and build on what we learn. The competition is open to everyone and free to enter. It will feature 3 tracks: - Track 1 is aimed at expert protein designers, with up to 20 teams to be selected. - Track 2 is targeting life science academics and industry researchers. - Track 3 is open to everyone from tech enthusiasts to high-school students. By combining Anthropic’s models with access to our automated lab, we want to make it possible for anyone with a laptop and an internet connection to join the global effort to advance human health with AI. A big thanks to Modal for contributing compute for protein design and to Twist Bioscience for contributing the DNA for the experimental validation! Sign up link below -
0:11

We’re partnering with Paul Jankura to launch the biggest Protein Design Competition in the world, challenging people around the world to use AI to design new potential drug candidates for diseases that affect millions of lives. The competition will feature five challenges, each focused on a specific disease or biological mechanism. Compared to previous competitions, it will be a big step-up in complexity and scale to push the boundaries of AI-driven protein design. Together with Anthropic, we’re sponsoring over $1 million in experimental validation, making it possible to test more than 5,000 protein designs in our automated lab at no cost to participants. Anthropic is providing an additional $1 million in Claude credits. All experimental results will be published openly on Proteinbase, including designs that didn’t work, so anyone can access the data and build on what we learn. The competition is open to everyone and free to enter. It will feature 3 tracks: - Track 1 is aimed at expert protein designers, with up to 20 teams to be selected. - Track 2 is targeting life science academics and industry researchers. - Track 3 is open to everyone from tech enthusiasts to high-school students. By combining Anthropic’s models with access to our automated lab, we want to make it possible for anyone with a laptop and an internet connection to join the global effort to advance human health with AI. A big thanks to Modal for contributing compute for protein design and to Twist Bioscience for contributing the DNA for the experimental validation! Sign up link below -

Adaptyv Bio

208,864 görüntüleme • 26 gün önce

"Europe has already lost the AI race." I hear this bashing almost every single week on LinkedIn and X. Not so much when I talk to the teams who are actually working on it. They are trying to do something about it. Take Feyer. They are not building another generic wrapper application. Feyer is developing AI systems that autonomously design novel industrial hardware. Their neural explorers are coupled directly with differentiable physics simulations, allowing them to search enormous design spaces and discover hardware that humans might never come up with themselves. That could accelerate innovation across everything from lasers and quantum technology to microchip production. On September 9, Cyber Valley celebrates its 10th anniversary here in Tübingen. And I want to show and talk about companies like Feyer that are sitting here. They are a pretty good example of what can happen when world-class research turns into an ambitious company. They are building right here in Tübingen and just secured €3 million in the SPRIND, Federal Agency for Breakthrough Innovation - Bundesagentur für Sprunginnovationen Next Frontier AI Challenge. I spent some time on a call with their CTO Sören Arlt last week, and the level of technical ambition there is exactly what this ecosystem needs right now. And Feyer is part of a much bigger bet. SPRIND, Federal Agency for Breakthrough Innovation is deploying €125 million over 24 months to build three internationally competitive European frontier AI labs. Ten teams start with up to €3 million each, six can advance with another €8 million, and the final three can receive another €15.5 million each. Up to €26.5 million per winning team. And this is not a research thesis invented for a startup pitch. The work builds on years of research by Sören Arlt, Mario Krenn and their collaborators into machine-driven scientific discovery. I am going to share a lot more about what Sören Arlt, Jonathan Klimesch, Mario Krenn and the rest of the Feyer team are building very soon. If you want to see what European frontier AI can actually look like, keep an eye on them. Follow for more insights into AI and robotics. {Quick animation by me for now. We’ll have much better visuals to share over the next few weeks ;)}
0:10

"Europe has already lost the AI race." I hear this bashing almost every single week on LinkedIn and X. Not so much when I talk to the teams who are actually working on it. They are trying to do something about it. Take Feyer. They are not building another generic wrapper application. Feyer is developing AI systems that autonomously design novel industrial hardware. Their neural explorers are coupled directly with differentiable physics simulations, allowing them to search enormous design spaces and discover hardware that humans might never come up with themselves. That could accelerate innovation across everything from lasers and quantum technology to microchip production. On September 9, Cyber Valley celebrates its 10th anniversary here in Tübingen. And I want to show and talk about companies like Feyer that are sitting here. They are a pretty good example of what can happen when world-class research turns into an ambitious company. They are building right here in Tübingen and just secured €3 million in the SPRIND, Federal Agency for Breakthrough Innovation - Bundesagentur für Sprunginnovationen Next Frontier AI Challenge. I spent some time on a call with their CTO Sören Arlt last week, and the level of technical ambition there is exactly what this ecosystem needs right now. And Feyer is part of a much bigger bet. SPRIND, Federal Agency for Breakthrough Innovation is deploying €125 million over 24 months to build three internationally competitive European frontier AI labs. Ten teams start with up to €3 million each, six can advance with another €8 million, and the final three can receive another €15.5 million each. Up to €26.5 million per winning team. And this is not a research thesis invented for a startup pitch. The work builds on years of research by Sören Arlt, Mario Krenn and their collaborators into machine-driven scientific discovery. I am going to share a lot more about what Sören Arlt, Jonathan Klimesch, Mario Krenn and the rest of the Feyer team are building very soon. If you want to see what European frontier AI can actually look like, keep an eye on them. Follow for more insights into AI and robotics. {Quick animation by me for now. We’ll have much better visuals to share over the next few weeks ;)}

Ilir Aliu

10,745 görüntüleme • 1 ay önce

🧵 Understanding Zama; the future of privacy tech & homomorphic encryption 1️⃣ Zama is pioneering fully homomorphic encryption (FHE). A breakthrough that lets you compute on encrypted data without decrypting it. 🔐 That means total privacy, even the system running your data can’t see it. 2️⃣ Why it matters: Right now, cloud apps, AI models, and databases must access your raw data to work. FHE changes that. your data stays private while still usable. 3️⃣ Zama builds open-source FHE tools for developers, turning advanced cryptography into practical products for AI, blockchain, and Web3. 4️⃣ Imagine: •AI that learns without reading your secrets 🤖 •Blockchain transactions with zero data leaks •Cloud apps that never see your info 5️⃣ Zama’s mission: Privacy should be the default, not an option. They’re making privacy-preserving tech simple, scalable, and open for everyone. 🔚 In a world obsessed with data, Zama might just be building the encryption layer of the future internet. 🌐
0:18

🧵 Understanding Zama; the future of privacy tech & homomorphic encryption 1️⃣ Zama is pioneering fully homomorphic encryption (FHE). A breakthrough that lets you compute on encrypted data without decrypting it. 🔐 That means total privacy, even the system running your data can’t see it. 2️⃣ Why it matters: Right now, cloud apps, AI models, and databases must access your raw data to work. FHE changes that. your data stays private while still usable. 3️⃣ Zama builds open-source FHE tools for developers, turning advanced cryptography into practical products for AI, blockchain, and Web3. 4️⃣ Imagine: •AI that learns without reading your secrets 🤖 •Blockchain transactions with zero data leaks •Cloud apps that never see your info 5️⃣ Zama’s mission: Privacy should be the default, not an option. They’re making privacy-preserving tech simple, scalable, and open for everyone. 🔚 In a world obsessed with data, Zama might just be building the encryption layer of the future internet. 🌐

v͙e͙s͙p͙e͙r͙ 📊🐐

19,644 görüntüleme • 11 ay önce

New open-source agent harness just landed! I got early access to TrueForge by TrueFoundry and have been running it locally for the past few days. The harness layer deserves as much attention as the model, and open source matters here because you can inspect the loop, run it on your own infrastructure, and swap to the latest or cheaper models. TrueForge handles the runtime work that makes an agent reliable. It drives the tool-calling loop, manages context, coordinates subagents, and executes code in a sandbox, with any model you choose. Every tool call re-sends the growing context to the model, so in practice the harness controls most of what an agent costs to run. A few things stood out from my testing and their published benchmarks. Vendor-Neutral by design. It runs OpenAI, Anthropic, and Google models alongside open-weight models like Kimi, GLM, and DeepSeek. Model routing is a setting, and you can send each task to the model that fits it. On a 14-task enterprise agent benchmark, it matched the accuracy of Claude Managed Agents running the same Opus 4.8 model at roughly 30% lower cost per run (3.8M tokens vs 10M for the same answers). Routing the same tasks to GLM-5.2 held accuracy and brought cost down by about 75%, around $3 per run instead of $12. Fully self-hosted and Open Source (MIT License). I had it running locally with one command, with sandboxed code execution working out of the box. It's time to own your agent harness. Thanks to TrueFoundry for partnering on this post.
0:16

New open-source agent harness just landed! I got early access to TrueForge by TrueFoundry and have been running it locally for the past few days. The harness layer deserves as much attention as the model, and open source matters here because you can inspect the loop, run it on your own infrastructure, and swap to the latest or cheaper models. TrueForge handles the runtime work that makes an agent reliable. It drives the tool-calling loop, manages context, coordinates subagents, and executes code in a sandbox, with any model you choose. Every tool call re-sends the growing context to the model, so in practice the harness controls most of what an agent costs to run. A few things stood out from my testing and their published benchmarks. Vendor-Neutral by design. It runs OpenAI, Anthropic, and Google models alongside open-weight models like Kimi, GLM, and DeepSeek. Model routing is a setting, and you can send each task to the model that fits it. On a 14-task enterprise agent benchmark, it matched the accuracy of Claude Managed Agents running the same Opus 4.8 model at roughly 30% lower cost per run (3.8M tokens vs 10M for the same answers). Routing the same tasks to GLM-5.2 held accuracy and brought cost down by about 75%, around $3 per run instead of $12. Fully self-hosted and Open Source (MIT License). I had it running locally with one command, with sandboxed code execution working out of the box. It's time to own your agent harness. Thanks to TrueFoundry for partnering on this post.

elvis

11,303 görüntüleme • 1 ay önce