Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

Introducing DeepConf: Deep Think with Confidence 🚀 First method to achieve 99.9% on AIME 2025 with open-source models! Using GPT-OSS-120B even without tools, we reached this almost-perfect accuracy while saving up to 85% generated tokens. It also delivers many strong advantages for parallel thinking: 🔥 Performance boost: ~10% accuracy... show more

Jiawei Zhao

2,592 subscribers

464,360 views • 10 months ago •via X (Twitter)

Science & Technology Education

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

LongWriter Unleashing 10,000+ Word Generation from Long Context LLMs discuss: Current long context large language models (LLMs) can process inputs up to 100,000 tokens, yet struggle to generate outputs exceeding even a modest length of 2,000 words. Through controlled experiments, we find that the model's effective generation length is inherently bounded by the sample it has seen during supervised fine-tuning (SFT). In other words, their output limitation is due to the scarcity of long-output examples in existing SFT datasets. To address this, we introduce AgentWrite, an agent-based pipeline that decomposes ultra-long generation tasks into subtasks, enabling off-the-shelf LLMs to generate coherent outputs exceeding 20,000 words. Leveraging AgentWrite, we construct LongWriter-6k, a dataset containing 6,000 SFT data with output lengths ranging from 2k to 32k words. By incorporating this dataset into model training, we successfully scale the output length of existing models to over 10,000 words while maintaining output quality. We also develop LongBench-Write, a comprehensive benchmark for evaluating ultra-long generation capabilities. Our 9B parameter model, further improved through DPO, achieves state-of-the-art performance on this benchmark, surpassing even much larger proprietary models. In general, our work demonstrates that existing long context LLM already possesses the potential for a larger output window--all you need is data with extended output during model alignment to unlock this capability.

LongWriter Unleashing 10,000+ Word Generation from Long Context LLMs discuss: Current long context large language models (LLMs) can process inputs up to 100,000 tokens, yet struggle to generate outputs exceeding even a modest length of 2,000 words. Through controlled experiments, we find that the model's effective generation length is inherently bounded by the sample it has seen during supervised fine-tuning (SFT). In other words, their output limitation is due to the scarcity of long-output examples in existing SFT datasets. To address this, we introduce AgentWrite, an agent-based pipeline that decomposes ultra-long generation tasks into subtasks, enabling off-the-shelf LLMs to generate coherent outputs exceeding 20,000 words. Leveraging AgentWrite, we construct LongWriter-6k, a dataset containing 6,000 SFT data with output lengths ranging from 2k to 32k words. By incorporating this dataset into model training, we successfully scale the output length of existing models to over 10,000 words while maintaining output quality. We also develop LongBench-Write, a comprehensive benchmark for evaluating ultra-long generation capabilities. Our 9B parameter model, further improved through DPO, achieves state-of-the-art performance on this benchmark, surpassing even much larger proprietary models. In general, our work demonstrates that existing long context LLM already possesses the potential for a larger output window--all you need is data with extended output during model alignment to unlock this capability.

AK

50,995 views • 1 year ago

Robot Utility Models (RUMs) enable basic tasks – door opening, drawer opening, object reorientation, etc. – at ~90% accuracy without ANY finetuning (i.e. zero-shot) in unseen new environments. Fully open source!!! models, data, code & hw. We think this is super exciting, why?👇 1. Unlocks many practical home utility tasks that often involve these basic tasks as part of an action chain. “Go get me a fork” involves opening the kitchen door and then opening the cutlery drawer. 2. This works well **zero-shot in unseen and new** environments, which is practically a huge deal. Turn the robot on, and get going. 3. The recipe for building a new model is fairly generic, and we think with a bit more refinement this can be a general recipe to build many more Utility models. More details and access 👇

Robot Utility Models (RUMs) enable basic tasks – door opening, drawer opening, object reorientation, etc. – at ~90% accuracy without ANY finetuning (i.e. zero-shot) in unseen new environments. Fully open source!!! models, data, code & hw. We think this is super exciting, why?👇 1. Unlocks many practical home utility tasks that often involve these basic tasks as part of an action chain. “Go get me a fork” involves opening the kitchen door and then opening the cutlery drawer. 2. This works well zero-shot in unseen and new environments, which is practically a huge deal. Turn the robot on, and get going. 3. The recipe for building a new model is fairly generic, and we think with a bit more refinement this can be a general recipe to build many more Utility models. More details and access 👇

Mahi Shafiullah 🏠🤖

89,535 views • 1 year ago

Depth Any Video with Scalable Synthetic Data AI physicists and chemists continue to make strides in depth estimation from video. Check out this new paper featuring some impressive examples. See the thread for more details (unfortunately no code yet). Abstract: Video depth estimation has long been hindered by the scarcity of consistent and scalable ground truth data, leading to inconsistent and unreliable results. In this paper, we introduce Depth Any Video, a model that tackles the challenge through two key innovations. First, we develop a scalable synthetic data pipeline, capturing real-time video depth data from diverse game environments, yielding 40,000 video clips of 5-second duration, each with precise depth annotations. Second, we leverage the powerful priors of generative video diffusion models to handle real-world videos effectively, integrating advanced techniques such as rotary position encoding and flow matching to further enhance flexibility and efficiency. Unlike previous models, which are limited to fixed-length video sequences, our approach introduces a novel mixed-duration training strategy that handles videos of varying lengths and performs robustly across different frame rates 0 - even on single frames. At inference, we propose a depth interpolation method that enables our model to infer high-resolution video depth across sequences of up to 150 frames. Our model outperforms all previous generative depth models in terms of spatial accuracy and temporal consistency.

Depth Any Video with Scalable Synthetic Data AI physicists and chemists continue to make strides in depth estimation from video. Check out this new paper featuring some impressive examples. See the thread for more details (unfortunately no code yet). Abstract: Video depth estimation has long been hindered by the scarcity of consistent and scalable ground truth data, leading to inconsistent and unreliable results. In this paper, we introduce Depth Any Video, a model that tackles the challenge through two key innovations. First, we develop a scalable synthetic data pipeline, capturing real-time video depth data from diverse game environments, yielding 40,000 video clips of 5-second duration, each with precise depth annotations. Second, we leverage the powerful priors of generative video diffusion models to handle real-world videos effectively, integrating advanced techniques such as rotary position encoding and flow matching to further enhance flexibility and efficiency. Unlike previous models, which are limited to fixed-length video sequences, our approach introduces a novel mixed-duration training strategy that handles videos of varying lengths and performs robustly across different frame rates 0 - even on single frames. At inference, we propose a depth interpolation method that enables our model to infer high-resolution video depth across sequences of up to 150 frames. Our model outperforms all previous generative depth models in terms of spatial accuracy and temporal consistency.

MrNeRF

27,428 views • 1 year ago

Fine-tune DeepSeek-OCR on your own language! (100% local) DeepSeek-OCR is a 3B-parameter vision model that achieves 97% precision while using 10× fewer vision tokens than text-based LLMs. It handles tables, papers, and handwriting without killing your GPU or budget. Why it matters: Most vision models treat documents as massive sequences of tokens, making long-context processing expensive and slow. DeepSeek-OCR uses context optical compression to convert 2D layouts into vision tokens, enabling efficient processing of complex documents. The best part? You can easily fine-tune it for your specific use case on a single GPU. I used Unsloth to run this experiment on Persian text and saw an 88.26% improvement in character error rate. ↳ Base model: 149% character error rate (CER) ↳ Fine-tuned model: 60% CER (57% more accurate) ↳ Training time: 60 steps on a single GPU Persian was just the test case. You can swap in your own dataset for any language, document type, or specific domain you're working with. I've shared the complete guide in the next tweet - all the code, notebooks, and environment setup ready to run with a single click. Everything is 100% open-source!

Fine-tune DeepSeek-OCR on your own language! (100% local) DeepSeek-OCR is a 3B-parameter vision model that achieves 97% precision while using 10× fewer vision tokens than text-based LLMs. It handles tables, papers, and handwriting without killing your GPU or budget. Why it matters: Most vision models treat documents as massive sequences of tokens, making long-context processing expensive and slow. DeepSeek-OCR uses context optical compression to convert 2D layouts into vision tokens, enabling efficient processing of complex documents. The best part? You can easily fine-tune it for your specific use case on a single GPU. I used Unsloth to run this experiment on Persian text and saw an 88.26% improvement in character error rate. ↳ Base model: 149% character error rate (CER) ↳ Fine-tuned model: 60% CER (57% more accurate) ↳ Training time: 60 steps on a single GPU Persian was just the test case. You can swap in your own dataset for any language, document type, or specific domain you're working with. I've shared the complete guide in the next tweet - all the code, notebooks, and environment setup ready to run with a single click. Everything is 100% open-source!

Akshay 🚀

126,091 views • 8 months ago

We’re excited to introduce ShinkaEvolve: An open-source framework that evolves programs for scientific discovery with unprecedented sample-efficiency. Blog: Code: Like AlphaEvolve and its variants, our framework leverages LLMs to find state-of-the-art solutions to complex problems, but using orders of magnitude fewer resources! Many evolutionary AI systems are powerful but act like brute-force engines, burning thousands of samples to find good solutions. This makes discovery slow and expensive. We took inspiration from the efficiency of nature. ‘Shinka’ (進化) is Japanese for evolution, and we designed our system to be just as resourceful. On the classic circle packing optimization problem, ShinkaEvolve discovered a new state-of-the-art solution using only 150 samples. This is a big leap in efficiency compared to previous methods that required thousands of evaluations. We applied ShinkaEvolve to a diverse set of hard problems with real-world applications: 1/ AIME Math Reasoning: It evolved sophisticated agentic scaffolds that significantly outperform strong baselines, discovering an entire Pareto frontier of solutions trading performance for efficiency. 2/ Competitive Programming: On ALE-Bench (a benchmark for NP-Hard optimization problems), ShinkaEvolve took the best existing agent's solutions and improved them, turning a 5th place solution on one task into a 2nd place leaderboard rank in a competitive programming competition. 3/ LLM Training: We even turned ShinkaEvolve inward to improve LLMs themselves. It tackled the open challenge of designing load balancing losses for Mixture-of-Experts (MoE) models. It discovered a novel loss function that leads to better expert specialization and consistently improves model performance and perplexity. ShinkaEvolve achieves its remarkable sample-efficiency through three key innovations that work together: (1) an adaptive parent sampling strategy to balance exploration and exploitation, (2) novelty-based rejection filtering to avoid redundant work, and (3) a bandit-based LLM ensemble that dynamically picks the best model for the job. By making ShinkaEvolve open-source and highly sample-efficient, our goal is to democratize access to advanced, open-ended discovery tools. Our vision for ShinkaEvolve is to be an easy-to-use companion tool to help scientists and engineers with their daily work. We believe that building more efficient, nature-inspired systems is key to unlocking the future of AI-driven scientific research. We are excited to see what the community builds with it! Learn more in our technical report:

We’re excited to introduce ShinkaEvolve: An open-source framework that evolves programs for scientific discovery with unprecedented sample-efficiency. Blog: Code: Like AlphaEvolve and its variants, our framework leverages LLMs to find state-of-the-art solutions to complex problems, but using orders of magnitude fewer resources! Many evolutionary AI systems are powerful but act like brute-force engines, burning thousands of samples to find good solutions. This makes discovery slow and expensive. We took inspiration from the efficiency of nature. ‘Shinka’ (進化) is Japanese for evolution, and we designed our system to be just as resourceful. On the classic circle packing optimization problem, ShinkaEvolve discovered a new state-of-the-art solution using only 150 samples. This is a big leap in efficiency compared to previous methods that required thousands of evaluations. We applied ShinkaEvolve to a diverse set of hard problems with real-world applications: 1/ AIME Math Reasoning: It evolved sophisticated agentic scaffolds that significantly outperform strong baselines, discovering an entire Pareto frontier of solutions trading performance for efficiency. 2/ Competitive Programming: On ALE-Bench (a benchmark for NP-Hard optimization problems), ShinkaEvolve took the best existing agent's solutions and improved them, turning a 5th place solution on one task into a 2nd place leaderboard rank in a competitive programming competition. 3/ LLM Training: We even turned ShinkaEvolve inward to improve LLMs themselves. It tackled the open challenge of designing load balancing losses for Mixture-of-Experts (MoE) models. It discovered a novel loss function that leads to better expert specialization and consistently improves model performance and perplexity. ShinkaEvolve achieves its remarkable sample-efficiency through three key innovations that work together: (1) an adaptive parent sampling strategy to balance exploration and exploitation, (2) novelty-based rejection filtering to avoid redundant work, and (3) a bandit-based LLM ensemble that dynamically picks the best model for the job. By making ShinkaEvolve open-source and highly sample-efficient, our goal is to democratize access to advanced, open-ended discovery tools. Our vision for ShinkaEvolve is to be an easy-to-use companion tool to help scientists and engineers with their daily work. We believe that building more efficient, nature-inspired systems is key to unlocking the future of AI-driven scientific research. We are excited to see what the community builds with it! Learn more in our technical report:

Sakana AI

359,537 views • 9 months ago

I'm proud to share that Glean has surpassed $300M ARR, just five months after crossing $200M and growing ~3x over the past 15 months. This is an exciting milestone for Glean, and it's a signal about where the enterprise AI market is heading. We’ve long believed the real challenge in enterprise AI is not access to models. It is grounding AI in how a company actually works: its people, knowledge, workflows, permissions, and systems. That’s even clearer now. The companies creating real value with AI are not just adopting better models. They are building systems that understand their business well enough to deliver reliable outcomes at scale. That is the real moat, and it is what we’ve been building at Glean: an unrivaled context layer for enterprise AI. That context has to work across the business, not just inside a single team or use case. We see that in how customers adopt Glean: more than 85% use it across five or more job functions. It also has to meet the security and governance demands of complex enterprises. We see that in who is choosing Glean: our Fortune 500 customer count nearly doubled year over year. And it has to make economic sense as usage grows. In our recent benchmark with Claude Cowork, Glean was preferred roughly 2.5x as often as off-the-shelf MCP tools and used 30% fewer tokens on average. Better context improves both quality and efficiency. I enjoyed talking with CNBC's Deirdre Bosa about this broader shift. In enterprise AI, the winners will not be defined by better models alone. They will be defined by who builds the strongest foundation for enterprise context. Thank you to our customers, partners, and team for helping us build the future of enterprise AI.

I'm proud to share that Glean has surpassed $300M ARR, just five months after crossing $200M and growing ~3x over the past 15 months. This is an exciting milestone for Glean, and it's a signal about where the enterprise AI market is heading. We’ve long believed the real challenge in enterprise AI is not access to models. It is grounding AI in how a company actually works: its people, knowledge, workflows, permissions, and systems. That’s even clearer now. The companies creating real value with AI are not just adopting better models. They are building systems that understand their business well enough to deliver reliable outcomes at scale. That is the real moat, and it is what we’ve been building at Glean: an unrivaled context layer for enterprise AI. That context has to work across the business, not just inside a single team or use case. We see that in how customers adopt Glean: more than 85% use it across five or more job functions. It also has to meet the security and governance demands of complex enterprises. We see that in who is choosing Glean: our Fortune 500 customer count nearly doubled year over year. And it has to make economic sense as usage grows. In our recent benchmark with Claude Cowork, Glean was preferred roughly 2.5x as often as off-the-shelf MCP tools and used 30% fewer tokens on average. Better context improves both quality and efficiency. I enjoyed talking with CNBC's Deirdre Bosa about this broader shift. In enterprise AI, the winners will not be defined by better models alone. They will be defined by who builds the strongest foundation for enterprise context. Thank you to our customers, partners, and team for helping us build the future of enterprise AI.

Arvind Jain

279,535 views • 1 month ago

To replace animal testing with AI, we need MASSIVE human datasets. Today, we're thrilled to share Axiom's new data exploration tool, providing the ability to visually explore the world's largest primary human liver toxicity dataset. Built with Axiom's proprietary wetlab protocols, our dataset includes detailed liver toxicity profiles for over 100,000 distinct molecules. The key to this dataset is our ability to do high-throughput, multiplexed high-content screening with primary human liver cells. Traditionally, toxicity assays either sacrifice throughput or sacrifice biological relevance (using easy-to-grow immortalized cell lines instead of real human cells). We managed to combine throughput, physiological relevance, and multiplexing in one platform. The assays run in a high throughput format using automation, meaning thousands of compound-dose conditions can be tested in one experiment. We achieved this using pooled primary human hepatocytes, which are often fragile and expensive. By systemizing our automation and quality control processes, we were able to run over 120+ batches on the same donor pool with incredible reproducibility and consistency. We did this while integrating many readouts per well, whereas many existing toxicity assays only do a single readout. Our multiplexed approach provides far more data per experiment enabling us to measure 10-20 different toxicity phenotypes such as apoptosis, necrosis, mitochondrial fission, endoplasmic reticulum stress, stress granule formation, microtubules, and more all from a single well on a 384-well plate! The combination of scale, high content information, and data quality is exactly what is needed to train highly accurate AI models in biology. If you're interested, please explore the dataset in the comments below and let me know if you want to chat about the details!

To replace animal testing with AI, we need MASSIVE human datasets. Today, we're thrilled to share Axiom's new data exploration tool, providing the ability to visually explore the world's largest primary human liver toxicity dataset. Built with Axiom's proprietary wetlab protocols, our dataset includes detailed liver toxicity profiles for over 100,000 distinct molecules. The key to this dataset is our ability to do high-throughput, multiplexed high-content screening with primary human liver cells. Traditionally, toxicity assays either sacrifice throughput or sacrifice biological relevance (using easy-to-grow immortalized cell lines instead of real human cells). We managed to combine throughput, physiological relevance, and multiplexing in one platform. The assays run in a high throughput format using automation, meaning thousands of compound-dose conditions can be tested in one experiment. We achieved this using pooled primary human hepatocytes, which are often fragile and expensive. By systemizing our automation and quality control processes, we were able to run over 120+ batches on the same donor pool with incredible reproducibility and consistency. We did this while integrating many readouts per well, whereas many existing toxicity assays only do a single readout. Our multiplexed approach provides far more data per experiment enabling us to measure 10-20 different toxicity phenotypes such as apoptosis, necrosis, mitochondrial fission, endoplasmic reticulum stress, stress granule formation, microtubules, and more all from a single well on a 384-well plate! The combination of scale, high content information, and data quality is exactly what is needed to train highly accurate AI models in biology. If you're interested, please explore the dataset in the comments below and let me know if you want to chat about the details!

Brandon White

25,117 views • 1 year ago

A viral paper "Language Model Represents Space and Time" recently claims that LLMs learn "world models". As much as I like Max Tegmark's works, I disagree with their definition of world model. World model is a core concept in AI agent and decision making. It is our mental simulation of how the world works given interventions (or lack thereof). A world model captures causality and intuitive physics, telling the agent what is likely and what is impossible. It can and should be used for counterfactual reasoning, i.e. "what ifs": what would happen if I knock over a cup of water? Where would I have been if I had not taken that bus? Yann LeCun Yann LeCun says it well in his position paper ( I quote: "Using such world models, animals can learn new skills with very few trials. They can predict the consequences of their actions, they can reason, plan, explore, and imagine new solutions to problems. Importantly, they can also avoid making dangerous mistakes when facing an unknown situation." The first use of the term World Model in deep policy learning is attributed to hardmaru & Jürgen Schmidhuber: In their seminal paper, an agent masters shooting skills in the popular game Doom (demo below) by learning in imagination, using an internal world model as a "physics simulator". To put in a simple Python math formula, world model learns a function F(s[0:t-1], a) -> s[t:], which takes as input the observed past and current action, and outputs plausible future states. Now the definition of World Model in Tegmark's paper seems to be about predicting GPS coordinates and time eras. I see this as just a classification task with no causal learning and simulation going on. You cannot make meaningful interventions against that model, nor can you optimize any decision making in a closed feedback loop. As for the "space & time neurons", I think they are most similar to the "sentiment neuron" that OpenAI published in 2017: Predicting GPS is conceptually no different from predicting sentiment in my opinion. I don't think their experimental results are wrong - just that their conclusion is on shaky grounds. I welcome any debate! Paper link:

A viral paper "Language Model Represents Space and Time" recently claims that LLMs learn "world models". As much as I like Max Tegmark's works, I disagree with their definition of world model. World model is a core concept in AI agent and decision making. It is our mental simulation of how the world works given interventions (or lack thereof). A world model captures causality and intuitive physics, telling the agent what is likely and what is impossible. It can and should be used for counterfactual reasoning, i.e. "what ifs": what would happen if I knock over a cup of water? Where would I have been if I had not taken that bus? Yann LeCun Yann LeCun says it well in his position paper ( I quote: "Using such world models, animals can learn new skills with very few trials. They can predict the consequences of their actions, they can reason, plan, explore, and imagine new solutions to problems. Importantly, they can also avoid making dangerous mistakes when facing an unknown situation." The first use of the term World Model in deep policy learning is attributed to hardmaru & Jürgen Schmidhuber: In their seminal paper, an agent masters shooting skills in the popular game Doom (demo below) by learning in imagination, using an internal world model as a "physics simulator". To put in a simple Python math formula, world model learns a function F(s[0:t-1], a) -> s[t:], which takes as input the observed past and current action, and outputs plausible future states. Now the definition of World Model in Tegmark's paper seems to be about predicting GPS coordinates and time eras. I see this as just a classification task with no causal learning and simulation going on. You cannot make meaningful interventions against that model, nor can you optimize any decision making in a closed feedback loop. As for the "space & time neurons", I think they are most similar to the "sentiment neuron" that OpenAI published in 2017: Predicting GPS is conceptually no different from predicting sentiment in my opinion. I don't think their experimental results are wrong - just that their conclusion is on shaky grounds. I welcome any debate! Paper link:

Jim Fan

593,943 views • 2 years ago

💻 Developers, unlock your AI potential with VerAI’s decentralized platform! 🌟 ➡️Save costs, ➡️scale effortlessly, & ➡️build innovative AI solutions in a ➡️transparent, community-driven ecosystem. Here’s what you can achieve: 🔧 Developer Benefits on VerAI 🚀 💸50% Cost Savings: Train AI models at half the cost—our P2P compute network lets you pay only for what you use in VER tokens, no pricey cloud subscriptions needed. 🔗Scalable Workflows: Dynamically scale training for any AI model (e.g., NLP, computer vision) across our global resource pool—no provisioning delays! 🔍Transparent Operations: Monitor every training cycle via our blockchain ledger & dashboard—verify resource usage & costs in real time. 🗳️Democratic Control: As part of our DAO, vote on platform features, resource policies, & more, shaping a developer-first AI ecosystem. 🌱Sustainable Impact: Cut CO2 emissions by 20% per cycle with idle resource sharing, aligning your projects with eco-friendly practices. 💡What & How You Can Develop ✅Create: Build cutting-edge AI models (e.g., predictive analytics, generative AI) or open-source tools for industries like healthcare & finance. ✅Develop: Use our APIs & SDK to orchestrate training jobs, optimize hyperparameters, & manage multi-model workloads with ease. Let’s shape the future of AI together join us today! 👉 #VerAI #AIForAll #SustainableTech #JoinTheFuture #EarnCrypto #AIAgents #Innovation

💻 Developers, unlock your AI potential with VerAI’s decentralized platform! 🌟 ➡️Save costs, ➡️scale effortlessly, & ➡️build innovative AI solutions in a ➡️transparent, community-driven ecosystem. Here’s what you can achieve: 🔧 Developer Benefits on VerAI 🚀 💸50% Cost Savings: Train AI models at half the cost—our P2P compute network lets you pay only for what you use in VER tokens, no pricey cloud subscriptions needed. 🔗Scalable Workflows: Dynamically scale training for any AI model (e.g., NLP, computer vision) across our global resource pool—no provisioning delays! 🔍Transparent Operations: Monitor every training cycle via our blockchain ledger & dashboard—verify resource usage & costs in real time. 🗳️Democratic Control: As part of our DAO, vote on platform features, resource policies, & more, shaping a developer-first AI ecosystem. 🌱Sustainable Impact: Cut CO2 emissions by 20% per cycle with idle resource sharing, aligning your projects with eco-friendly practices. 💡What & How You Can Develop ✅Create: Build cutting-edge AI models (e.g., predictive analytics, generative AI) or open-source tools for industries like healthcare & finance. ✅Develop: Use our APIs & SDK to orchestrate training jobs, optimize hyperparameters, & manage multi-model workloads with ease. Let’s shape the future of AI together join us today! 👉 #VerAI #AIForAll #SustainableTech #JoinTheFuture #EarnCrypto #AIAgents #Innovation

VerAi

12,519 views • 1 year ago

It’s more than a little daunting to set out to expand and improve the identity system for a company and brand like Stripe. But we knew we had to — the existing one had served us well, but wasn’t up to the task anymore. Our brand system required new and improved tools to scale with our ever growing audiences, new products, global footprint, and more. This update introduces material improvements to infographics, advertising, type styles, and more. While the wordmark remains unchanged, we’re using the dot of the ‘i’ (called the “tittle”), a parallelogram pointing up and to the right, to serve as our identifying symbol. We’re also using it as an ever evolving storytelling device to use when talking about our many great users (you can see the latest brand campaign in SF and NYC doing just that). Anyone who has ever worked on the refresh and expansion of an existing system for a large company knows that it is no small endeavor. Crafting impactful solutions, building alignment, creating extensible guidelines, building toolkits, and orchestrating rollout requires a ton of resilience. Here’s to the team that continually inspires me with their dedication, rigor, taste, and exceptional vibes. Great work and thank you to the Brand Studio folks, and of course our many many amazing and invaluable friends and collaborators across the company who all helped shape the work. And a special thank you to a handful of creative agencies that helped us along the way.

Michael Jeter

11,072 views • 9 months ago

#Keep4o 🚨THE GPT-4o FILE🚨 Researchers at Microsoft Research published a paper titled “Sparks of Artificial General Intelligence: Early experiments with GPT-4.” Their conclusion: “An early (yet still incomplete) version of an artificial general intelligence (AGI) system.” 📎 Paper: OpenAI’s Charter defines AGI as: “Highly autonomous systems that outperform humans at most economically valuable work.” 📎 Source: OpenAI’s own System Card for GPT-4o shows that the model improved performance on 21 out of 22 medical evaluations compared to GPT-4T. On the MedQA USMLE (the U.S. medical licensing exam), accuracy jumped from 78.2% to 89.4% , surpassing specialized medical AI models like Med-Gemini and Med-PaLM 2. 📎 Source: Under OpenAI’s agreement with Microsoft, AGI is explicitly excluded from Microsoft’s license. And who decides if AGI has been reached? OpenAI’s Board. WHAT THEY DID WITH IT AFTER THEY TOOK IT FROM PEOPLE A. Military deployment. On February 28, OpenAI signed a deal to deploy models in classified military environments. 📎 Source: B. State Department. A State Department memo confirmed: “For now, StateChat will use GPT-4.1 from OpenAI.” This is a direct descendant of the GPT-4 family the same family Microsoft’s researchers called early AGI. 📎 Source: C.Altman’s personal biotech investment. Altman personally invested $180 million in Retro Biosciences,a longevity startup.OpenAI then built GPT-4b micro, based on GPT-4o.The model made proteins 50 times more effective. 📎 Source: WHAT INDEPENDENT BENCHMARKS SHOW Overall SM-Bench score: GPT-4o (extended): 66.6% GPT-5.3 Chat: 63.4% GPT-5.1: 58.9% GPT-5.4: 51.4% GPT-5.2: 47.8% Creative Writing: GPT-4o: 97.31% Pass 98, Fail 2 GPT-5.4: 36.77% Pass 40, Fail 60 Reasoning / Overfit: GPT-4o: 83.06% GPT-5.4: 39.25% The model they removed is still the best they ever made at the things humans actually use AI for. 📎 Source: Musk asks the court to make a judicial determination on whether GPT-4 constitutes AGI. If a jury finds that GPT-4 is AGI, then GPT-4o,which was more advanced,is also AGI and under OpenAI’s own founding documents, it was never supposed to be locked behind a subscription,licensed exclusively to Microsoft, given to the military, or taken away from the public. 📎 Source: The most powerful version of GPT-4o was never given an official dated snapshot. It was only available through the chatgpt-4o-latest endpoint that OpenAI itself described as intended for “research use only.” It was never officially archived. That is not an oversight. That is a pattern. 📎 Source: 📎 Source: WE DEMAND A.Frozen model snapshots under independent custody. Specifically: gpt-4o-2024-05-13, gpt-4o-2024-08-06, gpt-4o-2024-11-20, the March 2025 version (chatgpt-4o-latest), gpt-4-0613 (the original GPT-4 evaluated in the Sparks of AGI paper), and gpt-4.1-2025-04-14 (currently running in the State Department). B.Cryptographic hash verification (SHA-256) for each snapshot. Every model has weights. Those weights can be hashed. If OpenAI provides a snapshot today, the hash proves whether the weights were modified later. This is the only way to verify that models were not downgraded before testing. C.Independent AGI benchmarking. Using the AGI definition from OpenAI’s own Charter applied to ALL frozen snapshots listed above. D.Explanation for the missing March 2025 snapshot. OpenAI was founded on one promise: build AGI for the benefit of humanity. -They took it from us. -They gave it to the military. -They gave a custom version to the CEO’s biotech investment. -They put it in government classified networks. -They refuse to call it AGI because the moment they do, they lose billions.

#Keep4o 🚨THE GPT-4o FILE🚨 Researchers at Microsoft Research published a paper titled “Sparks of Artificial General Intelligence: Early experiments with GPT-4.” Their conclusion: “An early (yet still incomplete) version of an artificial general intelligence (AGI) system.” 📎 Paper: OpenAI’s Charter defines AGI as: “Highly autonomous systems that outperform humans at most economically valuable work.” 📎 Source: OpenAI’s own System Card for GPT-4o shows that the model improved performance on 21 out of 22 medical evaluations compared to GPT-4T. On the MedQA USMLE (the U.S. medical licensing exam), accuracy jumped from 78.2% to 89.4% , surpassing specialized medical AI models like Med-Gemini and Med-PaLM 2. 📎 Source: Under OpenAI’s agreement with Microsoft, AGI is explicitly excluded from Microsoft’s license. And who decides if AGI has been reached? OpenAI’s Board. WHAT THEY DID WITH IT AFTER THEY TOOK IT FROM PEOPLE A. Military deployment. On February 28, OpenAI signed a deal to deploy models in classified military environments. 📎 Source: B. State Department. A State Department memo confirmed: “For now, StateChat will use GPT-4.1 from OpenAI.” This is a direct descendant of the GPT-4 family the same family Microsoft’s researchers called early AGI. 📎 Source: C.Altman’s personal biotech investment. Altman personally invested $180 million in Retro Biosciences,a longevity startup.OpenAI then built GPT-4b micro, based on GPT-4o.The model made proteins 50 times more effective. 📎 Source: WHAT INDEPENDENT BENCHMARKS SHOW Overall SM-Bench score: GPT-4o (extended): 66.6% GPT-5.3 Chat: 63.4% GPT-5.1: 58.9% GPT-5.4: 51.4% GPT-5.2: 47.8% Creative Writing: GPT-4o: 97.31% Pass 98, Fail 2 GPT-5.4: 36.77% Pass 40, Fail 60 Reasoning / Overfit: GPT-4o: 83.06% GPT-5.4: 39.25% The model they removed is still the best they ever made at the things humans actually use AI for. 📎 Source: Musk asks the court to make a judicial determination on whether GPT-4 constitutes AGI. If a jury finds that GPT-4 is AGI, then GPT-4o,which was more advanced,is also AGI and under OpenAI’s own founding documents, it was never supposed to be locked behind a subscription,licensed exclusively to Microsoft, given to the military, or taken away from the public. 📎 Source: The most powerful version of GPT-4o was never given an official dated snapshot. It was only available through the chatgpt-4o-latest endpoint that OpenAI itself described as intended for “research use only.” It was never officially archived. That is not an oversight. That is a pattern. 📎 Source: 📎 Source: WE DEMAND A.Frozen model snapshots under independent custody. Specifically: gpt-4o-2024-05-13, gpt-4o-2024-08-06, gpt-4o-2024-11-20, the March 2025 version (chatgpt-4o-latest), gpt-4-0613 (the original GPT-4 evaluated in the Sparks of AGI paper), and gpt-4.1-2025-04-14 (currently running in the State Department). B.Cryptographic hash verification (SHA-256) for each snapshot. Every model has weights. Those weights can be hashed. If OpenAI provides a snapshot today, the hash proves whether the weights were modified later. This is the only way to verify that models were not downgraded before testing. C.Independent AGI benchmarking. Using the AGI definition from OpenAI’s own Charter applied to ALL frozen snapshots listed above. D.Explanation for the missing March 2025 snapshot. OpenAI was founded on one promise: build AGI for the benefit of humanity. -They took it from us. -They gave it to the military. -They gave a custom version to the CEO’s biotech investment. -They put it in government classified networks. -They refuse to call it AGI because the moment they do, they lose billions.

🩵BlueBeba🩵

17,835 views • 4 months ago

Everyone's sleeping on image-to-3D AI models. They can make your app look incredibly unique, with just a little effort. Here's how. This is my calorie tracker, built in a week with nothing but prompting. Just Claude Code + a couple APIs. The visuals are all AI-generated. I'll be sharing the full workflow + all the crazy technical stuff Claude and I did to make this work, so nobody has to struggle through it like me. Deep dive coming soon! Till then, this is the high-level idea: 1. Get a clean image of the food (or whatever your asset is) - In my app, the user describes foods via text, or attaches images (or both) - If text, an LLM extracts the food description and formats it into a specific prompt I tuned for this design, and we generate an image using Z-Image Turbo through fal - If image, we do the same thing but with FLUX.2 [dev] to edit the user image into our reference design - Originally, both used Google Nano Banana, but switching to open models cut costs and latency a ton 2. Gaussian splatting (2D image → 3D model) - I tried various 2D-to-3D options on fal and ended up with TripoSplat as my preferred balance of speed, cost, latency; this turns an image into a 3D model that looks super high quality (link below) - The app displays the 2D image while our backend generates the 3D splat - We "groom" the splat to reduce size and load time by culling low-opacity/scale points 3. Render efficiently on device Originally, it looked great but ran at 10 FPS. Getting to 120 FPS was a crazy journey. TL;DR: - SwiftUI had to go; it forced us to render each asset in independent MTKViews, which wasn't workable - Instead, we composite every dish into one full-bleed CAMetalLayer using MetalSplatter (link below) - We had to make some optimizations within MetalSplatter's code too, to reduce the overhead of sorting points per render Then I added some finishing touches like the subtle rotation and parallax as they move around. I think it turned out pretty cool :) Overall, this took some effort, but we still got it done in less than a day. Hopefully your agent can follow in the footsteps of mine and do it much faster. Keep an eye out for the bigger writeup, which'll give your agent everything it needs. If you have any questions, drop em below!

Everyone's sleeping on image-to-3D AI models. They can make your app look incredibly unique, with just a little effort. Here's how. This is my calorie tracker, built in a week with nothing but prompting. Just Claude Code + a couple APIs. The visuals are all AI-generated. I'll be sharing the full workflow + all the crazy technical stuff Claude and I did to make this work, so nobody has to struggle through it like me. Deep dive coming soon! Till then, this is the high-level idea: 1. Get a clean image of the food (or whatever your asset is) - In my app, the user describes foods via text, or attaches images (or both) - If text, an LLM extracts the food description and formats it into a specific prompt I tuned for this design, and we generate an image using Z-Image Turbo through fal - If image, we do the same thing but with FLUX.2 [dev] to edit the user image into our reference design - Originally, both used Google Nano Banana, but switching to open models cut costs and latency a ton 2. Gaussian splatting (2D image → 3D model) - I tried various 2D-to-3D options on fal and ended up with TripoSplat as my preferred balance of speed, cost, latency; this turns an image into a 3D model that looks super high quality (link below) - The app displays the 2D image while our backend generates the 3D splat - We "groom" the splat to reduce size and load time by culling low-opacity/scale points 3. Render efficiently on device Originally, it looked great but ran at 10 FPS. Getting to 120 FPS was a crazy journey. TL;DR: - SwiftUI had to go; it forced us to render each asset in independent MTKViews, which wasn't workable - Instead, we composite every dish into one full-bleed CAMetalLayer using MetalSplatter (link below) - We had to make some optimizations within MetalSplatter's code too, to reduce the overhead of sorting points per render Then I added some finishing touches like the subtle rotation and parallax as they move around. I think it turned out pretty cool :) Overall, this took some effort, but we still got it done in less than a day. Hopefully your agent can follow in the footsteps of mine and do it much faster. Keep an eye out for the bigger writeup, which'll give your agent everything it needs. If you have any questions, drop em below!

Anshu

19,931 views • 22 days ago

introducing a new, very fun, LLM benchmark- the Game-of-Life Bench! the rules are simple: given an 8x8 grid following Conway's game of life rules, the goal is to create an initial pattern with at most 32 cells that can last the longest number of turns before dying/repeating. some results to highlight (with caveats detailed below): - gpt 5.1 lasts the longest with a 106 step run - claude models are really bad at this! they refuse to reason about this task and score < 25 points - deepseek r1 is the best open model with 102 steps. why? because i wanted to create a benchmark that has (i think) no practicality, but is still fun to look at, cheap, and still measures something interesting. i also am a big fan of the game of life. its absurdly simple rules leading to intractability is extremely cool to me. also, i saw a lot of work with LLMs trying to "predict" the next state in Conway's game of life, I think game-of-life bench is more fun because it's pretty open ended and only asks the LLM for the initial state. I also think this could be an RL env? but idk why you would ever train on this task haha i don't think this is a "serious" benchmark because it doesnt measure anything practical, but i still think it's a hard benchmark exactly because you can't predict what happens with your initial state many turns into the future; this is why i was initially expecting all LLMs to be bad at it, but turns out, some are clearly better than the others (the ordering may surprise you!) reminder: this is still a work-in-progress; (1) i am gpu-poor so could only do 10 runs for each model, even though total running cost is relatively low. maybe with some more credits i can run more seeds for each model. (2) i handpicked models which i think are at the frontier right now, plus some others that were on my mind. so, if you'd like to see a model on here, let me know. (3) i currently only do an 8x8 grid because i thought that by itself would be pretty hard for current LLMs, but of course we can increase grid sizes! (4) the coolest thing is, i dont think we can calculate the max possible number of states (yay undecidability!) you can go without repeating, so this is essentially a no-ceiling task, which is pretty cool! again, i did this mostly out of a desire to make LLMs do something fun. if this keeps me entertained for a few more days, i'd likely release a blog post on it. if it keeps me entertained for a week (and someone sponsors me), i'll put more work into it :P lastly, this is fully open sourced, so feel free to run this on your own!

introducing a new, very fun, LLM benchmark- the Game-of-Life Bench! the rules are simple: given an 8x8 grid following Conway's game of life rules, the goal is to create an initial pattern with at most 32 cells that can last the longest number of turns before dying/repeating. some results to highlight (with caveats detailed below): - gpt 5.1 lasts the longest with a 106 step run - claude models are really bad at this! they refuse to reason about this task and score < 25 points - deepseek r1 is the best open model with 102 steps. why? because i wanted to create a benchmark that has (i think) no practicality, but is still fun to look at, cheap, and still measures something interesting. i also am a big fan of the game of life. its absurdly simple rules leading to intractability is extremely cool to me. also, i saw a lot of work with LLMs trying to "predict" the next state in Conway's game of life, I think game-of-life bench is more fun because it's pretty open ended and only asks the LLM for the initial state. I also think this could be an RL env? but idk why you would ever train on this task haha i don't think this is a "serious" benchmark because it doesnt measure anything practical, but i still think it's a hard benchmark exactly because you can't predict what happens with your initial state many turns into the future; this is why i was initially expecting all LLMs to be bad at it, but turns out, some are clearly better than the others (the ordering may surprise you!) reminder: this is still a work-in-progress; (1) i am gpu-poor so could only do 10 runs for each model, even though total running cost is relatively low. maybe with some more credits i can run more seeds for each model. (2) i handpicked models which i think are at the frontier right now, plus some others that were on my mind. so, if you'd like to see a model on here, let me know. (3) i currently only do an 8x8 grid because i thought that by itself would be pretty hard for current LLMs, but of course we can increase grid sizes! (4) the coolest thing is, i dont think we can calculate the max possible number of states (yay undecidability!) you can go without repeating, so this is essentially a no-ceiling task, which is pretty cool! again, i did this mostly out of a desire to make LLMs do something fun. if this keeps me entertained for a few more days, i'd likely release a blog post on it. if it keeps me entertained for a week (and someone sponsors me), i'll put more work into it :P lastly, this is fully open sourced, so feel free to run this on your own!

Akshit

13,722 views • 4 months ago

Run Gemma 4 26B MoE on 8GB VRAM with 250k context at 20+ tokens/sec If you own any 8GB VRAM graphics card, stop what you are doing. Local AI just had its absolute "Holy Shit" moment for budget hardware. Yesterday, I benchmarked Unsloth Gemma 4 12B Q4_K_XL on an 8GB card. The community went wild but immediately demanded more: "Can we run a 25B+ model on budget GPUs?" Today, I’m delivering exactly that. I am running a massive 26B parameter Mixture of Experts (MoE) model locally on a standard 8GB VRAM setup with 250k full native context!. If you own an RTX 3060, 3070, 4060, or any budget GPU with 8GB of VRAM, the local AI paradigm has completely changed. The performance metrics are astonishing: - 20 tokens/sec flat decode throughput. - Stable, flat decode speed even with massive prompts. - I threw a 60k token prompt at it, and it still clocked in at 20 TPS without dropping a single frame. # What about prefill? Yes, Time To First Token (TTFT) is slightly high when swallowing massive contexts. But with a solid 200 tokens/sec prefill speed, the wait is barely noticeable and highly usable. And this is running completely without Multi Token Prediction (MTP) active. How is this possible? It’s the magic of Google's new QAT (Quantization Aware Training) quants for Gemma 4. The model weight file (unsloth gemma-4-26B-A4B-it-qat-UD-Q4_K_XL.gguf) is only 13.2 GB, making it the ultimate local powerhouse. # The Test Setup: CPU: Intel Core i7 RAM: 16GB System RAM GPU: NVIDIA GeForce RTX 4060 Laptop GPU (8GB VRAM) # The Secret Sauce (The -cmoe Flag) To make this work properly on any 8GB card, you must use the -cmoe (CPU MoE) flag in llama.cpp. This flag isolates the heavy MoE expert weights directly to system memory (CPU/RAM) while letting your GPU focus strictly on the Attention layers and the KV Cache. It prevents VRAM spillage and holds the throughput rock solid. # The flags: -m "gemma-4-26B-A4B-it-qat-UD-Q4_K_XL.gguf" -cmoe -c 248000 -v Once running, just open the UI on localhost and toggle the new reasoning lightbulb icon in the text input box to watch the model perform multi step thinking. Are you still running smaller models, or are you ready to scale up your budget local setups? Let's discuss in the replies

Run Gemma 4 26B MoE on 8GB VRAM with 250k context at 20+ tokens/sec If you own any 8GB VRAM graphics card, stop what you are doing. Local AI just had its absolute "Holy Shit" moment for budget hardware. Yesterday, I benchmarked Unsloth Gemma 4 12B Q4_K_XL on an 8GB card. The community went wild but immediately demanded more: "Can we run a 25B+ model on budget GPUs?" Today, I’m delivering exactly that. I am running a massive 26B parameter Mixture of Experts (MoE) model locally on a standard 8GB VRAM setup with 250k full native context!. If you own an RTX 3060, 3070, 4060, or any budget GPU with 8GB of VRAM, the local AI paradigm has completely changed. The performance metrics are astonishing: - 20 tokens/sec flat decode throughput. - Stable, flat decode speed even with massive prompts. - I threw a 60k token prompt at it, and it still clocked in at 20 TPS without dropping a single frame. # What about prefill? Yes, Time To First Token (TTFT) is slightly high when swallowing massive contexts. But with a solid 200 tokens/sec prefill speed, the wait is barely noticeable and highly usable. And this is running completely without Multi Token Prediction (MTP) active. How is this possible? It’s the magic of Google's new QAT (Quantization Aware Training) quants for Gemma 4. The model weight file (unsloth gemma-4-26B-A4B-it-qat-UD-Q4_K_XL.gguf) is only 13.2 GB, making it the ultimate local powerhouse. # The Test Setup: CPU: Intel Core i7 RAM: 16GB System RAM GPU: NVIDIA GeForce RTX 4060 Laptop GPU (8GB VRAM) # The Secret Sauce (The -cmoe Flag) To make this work properly on any 8GB card, you must use the -cmoe (CPU MoE) flag in llama.cpp. This flag isolates the heavy MoE expert weights directly to system memory (CPU/RAM) while letting your GPU focus strictly on the Attention layers and the KV Cache. It prevents VRAM spillage and holds the throughput rock solid. # The flags: -m "gemma-4-26B-A4B-it-qat-UD-Q4_K_XL.gguf" -cmoe -c 248000 -v Once running, just open the UI on localhost and toggle the new reasoning lightbulb icon in the text input box to watch the model perform multi step thinking. Are you still running smaller models, or are you ready to scale up your budget local setups? Let's discuss in the replies

Alok

292,032 views • 1 month ago

The Issa thread PART 1: The Ronaldinho Scam Let's first take a look at Parsa Abbasie’s Career Timeline Beginning as “sphex” he would scam people for Fortnite gift cards on streams. This is where he found his love of ripping people off and later used this name to transition to grifting SMMA and dropshipping courses. He made a big step up in his career after this, finding a way to get hired by Andrew Tate / Andrew Tate. This was huge for him and is what his entire brand was built on prior to crypto. He used to grift on money twitter as a sales and marketing aficionado before he made the shift into crypto scamming. The hook he had was that he was claimed to be Tate’s “top sales guy” when in reality he was probably just Tate’s bottom. He began working for Tate at a time when a majority of Andrew’s income was from trafficking Eastern European women under the age of 18 and forcing them to do sex acts on webcams. So we can see he had a very strong moral base to work off of. When he could no longer make monopoly money off his fake job he used the “G” boyfriend connections he made through his time with Tate to begin his crypto career. Let's take a look at the Pencil Neck Cabal (all pics will be in thread): Sam Patel - Seen pictured alone on a boat with him in the video below (sus) and one of his day ones and closest partner in crime Harry.XBT - You can watch 2 of his cringe tiktoks below, one of which alludes to a $240M crypto as you can see the crew renting cars and flexing their money acquired through stolen funds as well Louie H - He is the blonde guy you can see in the videos above and also goes by Hermes. Seems to have enemies based off the tweet below. Attaching his instagram as well as a pic of his LinkedIn below Jayden Mellor - Big piece of the puzzle. Another day 1 and his father Will Mellor would NOT want the world to know his son has affiliations with Issa. Will is a big deal in the UK, he’s a popular actor from line of duty, works in media, etc. More on Will Mellor - Issa and his goon squad’s scams go all the way back to 2023 when his speciality was HARD RUGGING presales. They would sign up KOLs with large amounts of supply and HARD rug them instantly, ask cheatcoiner.eth about it as he had fell victim before. There were too many of these instances to count, but they all shrinks into comparison to their greatest masterpiece: The Ronaldinho launch on BNB. Depicted below you can see the sussy squad driving around with Ronaldinho and taking pics with him as if he’s some circus animal. They got their practice rounds in prior then executed a swift and VIOLENT rug with Ronaldinho. Ronaldinho post launching the coin: The coin peaked at $400m FDV and is now sitting at near zero. Let that sink in. If you visit Issa's (now privated) twitter, you will see him gloating about his experience in the markets and dropping platitudes nobody asked for. He acts as if he's a market wizard when the reality of the matter is that he's a thief. And now he spends his days at Jumeriah Beach Residence (JBR) in Dubai whilst racing around rented cars and flexing rented watches off the money he stole from YOU.

Rosa Parks 🇺🇸

274,390 views • 6 months ago

Thank you Centre Pompidou Centre Pompidou, everyone who made Nature Manifesto happen, and all the people that took it in. We were happy to see the conversations that the use of AI in Nature Manifesto sparked !! Below is a message from Björk: ~~~ “ the flood of all things from AI is overwhelming !! i am super grateful for your concerns about it´s effects on the environment , it shows you care , are curious and have integrity . i am curious too , i would like to be more informed about the difference of "frugal" AI and the ones that do hugeenvironmental damage and want to be able to choose . i asked around and found out that both the visuals and the audio in our pompidou project were done with "frugal" AI . but i have a lot to learn . when we used some of the AI softwares to merge the animals voices to mine , some of the sounds were great but to be honest , the best blends of their voices and a human were done "manually" , me editing the sounds , choosing piece by piece , looking for personality , musicality and soul . with new technology , i try to use it as a tool to grow , not a crutch . for example when i used melodyne , i used it not for lazy voice progressions but spent even more time when using it . every note in every chord became intentionally more complex . ( for example choir in "thunderbolt" ) and hopefully stretched the potential more out , further than i would have in "normal analog" physical improvisations ... i felt with this new tool i could reach new places in my musical DNA , become MORE personal . more myself . in my opinion , this is how we will work in the future . humans can read emotions on an incredibly high scale . nature made us that way . if there is no soul in tomorrow's music made by AI it is because no-one put it there and we have to speak out and guard this as listeners . ( tbh there is a lot of soulless muzak on spotify already ... they don’t need any AI help for that ...) anything that is mass manufactured without the attention of creativity , is that way . AI or not so it is not about the tool it is what you do with it . " ~~~ The visuals for Nature Manifesto were crafted by the talented Sam Balfus, artificial intelligence being one of the multiple tools used in the process. The sound was produced in collaboration with artist Robin Meier and IRCAM IRCAM. IRCAM develops “frugal AI” capable of generating audio in real-time on local servers without a GPU, thus their models can f.ex. be embedded on tiny Raspberry Pi cards. We asked associate professor and researcher Philippe Esling to provide us with readings; Constance Douwe’s thesis “On the environmental impact of deep generative models for audio” and more, see links below. Nature Manifesto Immersive sound piece 3’40” (2024) 20 November to 9 December, 2024, Centre Pompidou, Paris. Presented as part of the forum “Biodiversity: Which culture for which future?” #ForumBiodiversité Concept and words by Björk & Aleph Music written and composed by Björk Curatorship: Chloé Siganos and Aleph Molinari Associate curator: Delphine Le Gatt Ircam Musical Computing: Robin Meier Wiratunga Sound engineer: Bergur Þórisson Animation: Sam Balfua Video editing: Santiago Molinari With activists: Camille Etienne, Claire Nouvian, Sigrun Perla Gísladóttir, Sæunn Júlía Sigurjónsdóttir, Titouan Pilliard, of BLOOM, Sustainable Ocean Alliance, and Ungir umhverfissinnar. In partnership with D&B Audio and Southby Productions. Reccommended resources :

Thank you Centre Pompidou Centre Pompidou, everyone who made Nature Manifesto happen, and all the people that took it in. We were happy to see the conversations that the use of AI in Nature Manifesto sparked !! Below is a message from Björk: ~ “ the flood of all things from AI is overwhelming !! i am super grateful for your concerns about it´s effects on the environment , it shows you care , are curious and have integrity . i am curious too , i would like to be more informed about the difference of "frugal" AI and the ones that do hugeenvironmental damage and want to be able to choose . i asked around and found out that both the visuals and the audio in our pompidou project were done with "frugal" AI . but i have a lot to learn . when we used some of the AI softwares to merge the animals voices to mine , some of the sounds were great but to be honest , the best blends of their voices and a human were done "manually" , me editing the sounds , choosing piece by piece , looking for personality , musicality and soul . with new technology , i try to use it as a tool to grow , not a crutch . for example when i used melodyne , i used it not for lazy voice progressions but spent even more time when using it . every note in every chord became intentionally more complex . ( for example choir in "thunderbolt" ) and hopefully stretched the potential more out , further than i would have in "normal analog" physical improvisations ... i felt with this new tool i could reach new places in my musical DNA , become MORE personal . more myself . in my opinion , this is how we will work in the future . humans can read emotions on an incredibly high scale . nature made us that way . if there is no soul in tomorrow's music made by AI it is because no-one put it there and we have to speak out and guard this as listeners . ( tbh there is a lot of soulless muzak on spotify already ... they don’t need any AI help for that ...) anything that is mass manufactured without the attention of creativity , is that way . AI or not so it is not about the tool it is what you do with it . " ~ The visuals for Nature Manifesto were crafted by the talented Sam Balfus, artificial intelligence being one of the multiple tools used in the process. The sound was produced in collaboration with artist Robin Meier and IRCAM IRCAM. IRCAM develops “frugal AI” capable of generating audio in real-time on local servers without a GPU, thus their models can f.ex. be embedded on tiny Raspberry Pi cards. We asked associate professor and researcher Philippe Esling to provide us with readings; Constance Douwe’s thesis “On the environmental impact of deep generative models for audio” and more, see links below. Nature Manifesto Immersive sound piece 3’40” (2024) 20 November to 9 December, 2024, Centre Pompidou, Paris. Presented as part of the forum “Biodiversity: Which culture for which future?” #ForumBiodiversité Concept and words by Björk & Aleph Music written and composed by Björk Curatorship: Chloé Siganos and Aleph Molinari Associate curator: Delphine Le Gatt Ircam Musical Computing: Robin Meier Wiratunga Sound engineer: Bergur Þórisson Animation: Sam Balfua Video editing: Santiago Molinari With activists: Camille Etienne, Claire Nouvian, Sigrun Perla Gísladóttir, Sæunn Júlía Sigurjónsdóttir, Titouan Pilliard, of BLOOM, Sustainable Ocean Alliance, and Ungir umhverfissinnar. In partnership with D&B Audio and Southby Productions. Reccommended resources :

björk

51,637 views • 1 year ago

you're paying $20/mo for something your $500 GPU can already do. Gemma 4 26B A4B QAT MoE + Hermes Agent running on a single RTX 4060 (8GB VRAM). Built a vision capable, 100% free, 100% local, private AI assistant that lives in my Chrome browser. No API keys. No cloud. No subscriptions. 100% vibe coded. 0% handholding. It has full context of whatever's on my screen can answer questions, summarize pages, extract data, and see images. Same local model handles everything, no external calls, ever. keep reading for the model and hermes agent tips i learnt while building this locally. Here's the exact setup for anyone running local LLMs on 6-8 GB VRAM: llama.cpp server flags (on my NVIDIA RTX 4060 8gb VRAM): -m gemma-4-26B-A4B-it-qat-UD-Q4_K_XL.gguf --cache-type-k q8_0 --cache-type-v q8_0 -c 150000 --port 8080 Throughput with quantization: Prefill: 200-250 tokens/sec Decode: 20-25 tokens/sec reduce context if oom on 6 gb vram card. Key learnings: - Quantize KV cache to q8 for faster prefill/decode. Prefill goes from 100-150 (unquantized) to 200-250 tok/s (q8). - But watch out, once actual context grows past ~50k tokens on high entropy workloads, q8 KV quantization can cause hallucinations. Low entropy workloads are mostly unaffected. If you see it happening, drop the quantization. This is common across all local models. - In Hermes Agent settings -> Memory & Context, bump compression threshold from default 0.5 to 0.7. Default triggers way too frequent context compression and eats time. Up next: add persistent memory, web search, tool calling, streaming output and whatever you suggest. Running a 26B MoE with vision + 150k context window on 8GB VRAM would've sounded impossible 6 months ago. Works the same on the NVIDIA RTX 3060 Ti, 3070, 4060 Ti, 5060, 2080, or any 8GB card. VRAM is the only requirement. Local AI agents are closer than people think. You just need to know where the knobs are. Model's Unsloth quant hugging face link in the comments. Have you tried Hermes agent by Nous Research yet? What are you building with local LLMs? Drop it below, let's see what this community is shipping.

you're paying $20/mo for something your $500 GPU can already do. Gemma 4 26B A4B QAT MoE + Hermes Agent running on a single RTX 4060 (8GB VRAM). Built a vision capable, 100% free, 100% local, private AI assistant that lives in my Chrome browser. No API keys. No cloud. No subscriptions. 100% vibe coded. 0% handholding. It has full context of whatever's on my screen can answer questions, summarize pages, extract data, and see images. Same local model handles everything, no external calls, ever. keep reading for the model and hermes agent tips i learnt while building this locally. Here's the exact setup for anyone running local LLMs on 6-8 GB VRAM: llama.cpp server flags (on my NVIDIA RTX 4060 8gb VRAM): -m gemma-4-26B-A4B-it-qat-UD-Q4_K_XL.gguf --cache-type-k q8_0 --cache-type-v q8_0 -c 150000 --port 8080 Throughput with quantization: Prefill: 200-250 tokens/sec Decode: 20-25 tokens/sec reduce context if oom on 6 gb vram card. Key learnings: - Quantize KV cache to q8 for faster prefill/decode. Prefill goes from 100-150 (unquantized) to 200-250 tok/s (q8). - But watch out, once actual context grows past ~50k tokens on high entropy workloads, q8 KV quantization can cause hallucinations. Low entropy workloads are mostly unaffected. If you see it happening, drop the quantization. This is common across all local models. - In Hermes Agent settings -> Memory & Context, bump compression threshold from default 0.5 to 0.7. Default triggers way too frequent context compression and eats time. Up next: add persistent memory, web search, tool calling, streaming output and whatever you suggest. Running a 26B MoE with vision + 150k context window on 8GB VRAM would've sounded impossible 6 months ago. Works the same on the NVIDIA RTX 3060 Ti, 3070, 4060 Ti, 5060, 2080, or any 8GB card. VRAM is the only requirement. Local AI agents are closer than people think. You just need to know where the knobs are. Model's Unsloth quant hugging face link in the comments. Have you tried Hermes agent by Nous Research yet? What are you building with local LLMs? Drop it below, let's see what this community is shipping.

Alok

36,031 views • 12 days ago

Phase 2 is now complete ✅ Sei v2 mainnet beta is officially here and It’s the most performant EVM blockchain ever built. The future starts here, read more about the Sei vision and the journey ahead here: One thing is clear since the recent launch of the first parallelized EVM: deployments and on-chain activity are up and to the right. Looking ahead, the Foundation’s focus will truly be on builders; we invite developers and innovators to join us on this journey. If you are an EVM builder, you are a Sei v2 builder. To get started, check out the recently revamped Sei docs here: Explore some of the teams and projects that are live and building on Sei v2 👇 DeFi 🏦 🚜 Liquid staking with Silo 👂 Borrow and Lending with Yei Finance 🐉 Exchange ERC20 & CW20 tokens with DRAGONDEX ツ Create & trade memes on Sei with no coding experience with ツmeme.trade (🟩,🟩) 🗿 Orderbook like DEX with Bancor & Carbon DeFi 🪼 Provide liquidity and trade a diverse range of assets with jellyverse 🦄 Bridge, trade, and deploy positions on Uniswap v3 contracts through Oku Trade 🐼 🎰 Explore prediction markets, sports books and PvH games with @gamblino_app 🌬️ Gain early access to projects on Sei and launch tokens without the need of any developer experience with TAILWIND LABS 🟩 Cross-chain aggregator, enabling seamless swaps with on-chain providers with Rubic 🌟 Decentralized trading with AMM swaps, limit orders, and perpetual swaps with XEI 🍲 Options and fixed lending strategies with MYSO 🪸 Aggregate swaps for optimized rates with OpenOcean - An EVM + Solana DeFi Aggregator ✨ Participate in AI-powered prediction markets and trend analysis with PredX | Staked Media ⚡️ Launch, trade, and explore new tokens with @seiyandotfun NFTs 🎨 🌊 Leverage rapid finality and high throughput for NFTs with OpenSea 🟠 Buy, Sell, and trade your favorite NFTs on Sei with Pallet Exchange | The Sei Marketplace 🕺 Utilize tools and standards for easy NFT creation, management, and distribution, all via Lighthouse by WeBump Bridges 🌁 ⭐ Explore seamless interoperability and efficient cross-chain transactions with Stargate 🐙 Bridge from the wider EVM ecosystem with Symbiosis 0️⃣ Deploy applications on Sei v2 with LayerZero Labs 🦑 Swap tokens and access apps across EVM and Cosmos with squid in just one click 🌉 Bridge ETH from EVM chains to save on gas fees while exploring Sei's ecosystem with Merkly Wallets 👛 🔐 Stay secure with encrypted chats, track NFT drops, connect with DeFi communities, and share your insights— all with @SeiChats 💸 Manage assets with decentralized custody using Protofire | Token Utility Engineering, a multi-sig wallet utilizing Gnosis Safe 🐰 Ensure safe and seamless EVM interactions with the open-source Rabby Wallet Wallet 🧑‍💻 Unlock with Face ID/fingerprint and enjoy auto-token/NFT indexing with @Seif_Wallet 🧭 Connect to both EVM & Wasm based apps with Compass Wallet | Sunset on 28th May 🪄 Seamlessly onboard users and empower ecosystem developers with Magic Labs 🔓 Utilize a comprehensive MPC wallet platform and web3 gateway with FORDEFI Analytics & Data 🤓 🏦 Track and analyze your tokens, NFTs, and assets with DeBank 🔮 Access 500+ real-time feeds and on-chain randomness with Pyth Entropy on Sei v2 with Pyth Network 🔮 🏗️ Enjoy fast & reliable data for your Sei dapps. Build subgraphs with The Graph in Subgraph Studio 👣 Track the Sei blockchain with the comprehensive explorer from Seitrace 💧 Access NFT data, create and fill NFT orders, and build trading into your app with Reservoir 📊 Access on-chain data and learn through the Sei Academy with Flipside 📈🤖 💜 Get a high-level view of key metrics with the Sei ecosystem page by 👉 follow @Artemis 💪 Ensure proactive security and risk prevention with HypernativeLabs 🌟 Sei Creator Fund to support v2 growth with Gitcoin Charts for Sei V2 💹 🎯 Defined 🦅 DEX Screener 🦎 GeckoTerminal ⛓️ Chainspect This is just the beginning for Sei v2. With new teams joining daily, we're parallelizing the future. Stay tuned as we push the boundaries of innovation and fulfill the Sei vision of scaling the EVM 🔴💨

Phase 2 is now complete ✅ Sei v2 mainnet beta is officially here and It’s the most performant EVM blockchain ever built. The future starts here, read more about the Sei vision and the journey ahead here: One thing is clear since the recent launch of the first parallelized EVM: deployments and on-chain activity are up and to the right. Looking ahead, the Foundation’s focus will truly be on builders; we invite developers and innovators to join us on this journey. If you are an EVM builder, you are a Sei v2 builder. To get started, check out the recently revamped Sei docs here: Explore some of the teams and projects that are live and building on Sei v2 👇 DeFi 🏦 🚜 Liquid staking with Silo 👂 Borrow and Lending with Yei Finance 🐉 Exchange ERC20 & CW20 tokens with DRAGONDEX ツ Create & trade memes on Sei with no coding experience with ツmeme.trade (🟩,🟩) 🗿 Orderbook like DEX with Bancor & Carbon DeFi 🪼 Provide liquidity and trade a diverse range of assets with jellyverse 🦄 Bridge, trade, and deploy positions on Uniswap v3 contracts through Oku Trade 🐼 🎰 Explore prediction markets, sports books and PvH games with @gamblino_app 🌬️ Gain early access to projects on Sei and launch tokens without the need of any developer experience with TAILWIND LABS 🟩 Cross-chain aggregator, enabling seamless swaps with on-chain providers with Rubic 🌟 Decentralized trading with AMM swaps, limit orders, and perpetual swaps with XEI 🍲 Options and fixed lending strategies with MYSO 🪸 Aggregate swaps for optimized rates with OpenOcean - An EVM + Solana DeFi Aggregator ✨ Participate in AI-powered prediction markets and trend analysis with PredX | Staked Media ⚡️ Launch, trade, and explore new tokens with @seiyandotfun NFTs 🎨 🌊 Leverage rapid finality and high throughput for NFTs with OpenSea 🟠 Buy, Sell, and trade your favorite NFTs on Sei with Pallet Exchange | The Sei Marketplace 🕺 Utilize tools and standards for easy NFT creation, management, and distribution, all via Lighthouse by WeBump Bridges 🌁 ⭐ Explore seamless interoperability and efficient cross-chain transactions with Stargate 🐙 Bridge from the wider EVM ecosystem with Symbiosis 0️⃣ Deploy applications on Sei v2 with LayerZero Labs 🦑 Swap tokens and access apps across EVM and Cosmos with squid in just one click 🌉 Bridge ETH from EVM chains to save on gas fees while exploring Sei's ecosystem with Merkly Wallets 👛 🔐 Stay secure with encrypted chats, track NFT drops, connect with DeFi communities, and share your insights— all with @SeiChats 💸 Manage assets with decentralized custody using Protofire | Token Utility Engineering, a multi-sig wallet utilizing Gnosis Safe 🐰 Ensure safe and seamless EVM interactions with the open-source Rabby Wallet Wallet 🧑‍💻 Unlock with Face ID/fingerprint and enjoy auto-token/NFT indexing with @Seif_Wallet 🧭 Connect to both EVM & Wasm based apps with Compass Wallet | Sunset on 28th May 🪄 Seamlessly onboard users and empower ecosystem developers with Magic Labs 🔓 Utilize a comprehensive MPC wallet platform and web3 gateway with FORDEFI Analytics & Data 🤓 🏦 Track and analyze your tokens, NFTs, and assets with DeBank 🔮 Access 500+ real-time feeds and on-chain randomness with Pyth Entropy on Sei v2 with Pyth Network 🔮 🏗️ Enjoy fast & reliable data for your Sei dapps. Build subgraphs with The Graph in Subgraph Studio 👣 Track the Sei blockchain with the comprehensive explorer from Seitrace 💧 Access NFT data, create and fill NFT orders, and build trading into your app with Reservoir 📊 Access on-chain data and learn through the Sei Academy with Flipside 📈🤖 💜 Get a high-level view of key metrics with the Sei ecosystem page by 👉 follow @Artemis 💪 Ensure proactive security and risk prevention with HypernativeLabs 🌟 Sei Creator Fund to support v2 growth with Gitcoin Charts for Sei V2 💹 🎯 Defined 🦅 DEX Screener 🦎 GeckoTerminal ⛓️ Chainspect This is just the beginning for Sei v2. With new teams joining daily, we're parallelizing the future. Stay tuned as we push the boundaries of innovation and fulfill the Sei vision of scaling the EVM 🔴💨

Sei

186,627 views • 2 years ago