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.

1000 FPS in a full 24/24 Deathmatch 🤯 Fully optimized system (Windows + BIOS tuned the right way) Consistent performance even in a full lobby Specs: • i7 13700KF • B760M Aorus Elite DDR4 • 32GB 3200MHz CL16 (16x2) • RTX 3070 Ti ⚡

honest | PC Tweaker

4,197 subscribers

108,389 views • 4 months ago •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

🤦‍♂️ James O’Brien has officially lost the plot. He’s just announced, with full seriousness, that he is now “feeling climate change in my stomach.” Not in the data. Not in the graphs. Not even in the weather. He just knows. His stomach says so. 🙃 Quote: “I am now feeling climate change in my stomach in ways that I don’t think I was doing until this summer. I’m embarrassed by that, but not enormously embarrassed by it. I can’t control it.” He even admits he used to listen to callers having full emotional meltdowns and think: “Why are you feeling that so keenly? I know you’re right… but I’m just not feeling it the same way.” Well done James. The climate gods have finally blessed your digestive system. You’ve levelled up from concerned broadcaster to man whose stomach is now a climate sensor. If this isn’t peak performance art, I don’t know what is. Next week: “I’m experiencing net zero in my left knee.” 😂

🤦‍♂️ James O’Brien has officially lost the plot. He’s just announced, with full seriousness, that he is now “feeling climate change in my stomach.” Not in the data. Not in the graphs. Not even in the weather. He just knows. His stomach says so. 🙃 Quote: “I am now feeling climate change in my stomach in ways that I don’t think I was doing until this summer. I’m embarrassed by that, but not enormously embarrassed by it. I can’t control it.” He even admits he used to listen to callers having full emotional meltdowns and think: “Why are you feeling that so keenly? I know you’re right… but I’m just not feeling it the same way.” Well done James. The climate gods have finally blessed your digestive system. You’ve levelled up from concerned broadcaster to man whose stomach is now a climate sensor. If this isn’t peak performance art, I don’t know what is. Next week: “I’m experiencing net zero in my left knee.” 😂

J Stewart

22,142 views • 7 days ago

OpenClaw setup made me $23,472 Literally overnight my $100 turned into $2,411 Average bot win rate 71% Copytrade: Here is the full strategy: The system builds automated workflows for trading by turning domain expertise into structured skills that activate automatically when specific market conditions appear Skill architecture Each skill is a modular package that includes instruction scripts and reference data This allows the system to apply specialized workflows without needing manual input for every trade Progressive context loading Skills use a three layer structure Only minimal metadata loads at first Full instructions historical data and supporting resources load only when required This reduces resource usage while keeping advanced trading capability Trigger detection Skills activate automatically when market conditions match predefined triggers such as volatility levels orderflow behavior or news sentiment This ensures the right workflow is used at the right time without manual action Workflow execution Once activated each skill runs a predefined multi step process including Real time price tracking and order book analysis Factor generation and backtesting Signal aggregation from machine learning models news sentiment and orderflow Risk assessment and capital allocation Trade execution with retries and position splitting Consistency and reliability All workflows are embedded directly into the system which ensures consistent execution instead of random decision making Every factor signal and risk rule is applied in a structured way Testing and iteration Skills are continuously improved using historical backtesting simulated trading and live performance tracking to maintain reliability in real market conditions Automation edge Instead of creating new strategies every time the system repeatedly uses optimized workflows This reduces complexity increases consistency and scales performance across thousands of trades Performance snapshot Started one month ago with $500 Current daily profit $2,300 per day Morning profit today $71,452 The system runs fully autonomously constantly scanning markets generating signals auditing trades managing risk and executing orders to maximize compounding returns

OpenClaw setup made me $23,472 Literally overnight my $100 turned into $2,411 Average bot win rate 71% Copytrade: Here is the full strategy: The system builds automated workflows for trading by turning domain expertise into structured skills that activate automatically when specific market conditions appear Skill architecture Each skill is a modular package that includes instruction scripts and reference data This allows the system to apply specialized workflows without needing manual input for every trade Progressive context loading Skills use a three layer structure Only minimal metadata loads at first Full instructions historical data and supporting resources load only when required This reduces resource usage while keeping advanced trading capability Trigger detection Skills activate automatically when market conditions match predefined triggers such as volatility levels orderflow behavior or news sentiment This ensures the right workflow is used at the right time without manual action Workflow execution Once activated each skill runs a predefined multi step process including Real time price tracking and order book analysis Factor generation and backtesting Signal aggregation from machine learning models news sentiment and orderflow Risk assessment and capital allocation Trade execution with retries and position splitting Consistency and reliability All workflows are embedded directly into the system which ensures consistent execution instead of random decision making Every factor signal and risk rule is applied in a structured way Testing and iteration Skills are continuously improved using historical backtesting simulated trading and live performance tracking to maintain reliability in real market conditions Automation edge Instead of creating new strategies every time the system repeatedly uses optimized workflows This reduces complexity increases consistency and scales performance across thousands of trades Performance snapshot Started one month ago with $500 Current daily profit $2,300 per day Morning profit today $71,452 The system runs fully autonomously constantly scanning markets generating signals auditing trades managing risk and executing orders to maximize compounding returns

winkle.

44,158 views • 5 months ago

My bet: in the near future, 80%⬆️ of CS research will be done by AI in collaboration with humans. However, today's research ecosystem is still built around the human, not the AI scientist. For example, the 8-page paper PDF is a lossy compression of months of branching exploration into a linear story, optimized for a human reviewer to skim in 30 minutes. It hides two structural taxes: 📖 Storytelling Tax — failures, rejected hypotheses, and dead ends get stripped. On RE-Bench (24,008 runs, 21 frontier models), failed runs = 90.2% of total compute cost, with a 113× median failed-to-success token ratio. Every lab independently rediscovers the same dead ends. 🔧 Engineering Tax — the gap between reviewer-sufficient prose and agent-sufficient spec. Across 8,921 PaperBench requirements (23 ICML'24 papers), only 45.4% are fully specified in the PDF. The rest is tacit lab knowledge. Tolerable when readers were human. Critical now that agents read, reproduce, and extend. We propose ARA: the Agent-Native Research Artifact — replace the narrative PDF with an agent-executable package, in 4 layers: 🧠 structured scientific logic ⚙️ executable code w/ full specs 🌳 exploration graph (every failure preserved) 📊 evidence grounding every claim

My bet: in the near future, 80%⬆️ of CS research will be done by AI in collaboration with humans. However, today's research ecosystem is still built around the human, not the AI scientist. For example, the 8-page paper PDF is a lossy compression of months of branching exploration into a linear story, optimized for a human reviewer to skim in 30 minutes. It hides two structural taxes: 📖 Storytelling Tax — failures, rejected hypotheses, and dead ends get stripped. On RE-Bench (24,008 runs, 21 frontier models), failed runs = 90.2% of total compute cost, with a 113× median failed-to-success token ratio. Every lab independently rediscovers the same dead ends. 🔧 Engineering Tax — the gap between reviewer-sufficient prose and agent-sufficient spec. Across 8,921 PaperBench requirements (23 ICML'24 papers), only 45.4% are fully specified in the PDF. The rest is tacit lab knowledge. Tolerable when readers were human. Critical now that agents read, reproduce, and extend. We propose ARA: the Agent-Native Research Artifact — replace the narrative PDF with an agent-executable package, in 4 layers: 🧠 structured scientific logic ⚙️ executable code w/ full specs 🌳 exploration graph (every failure preserved) 📊 evidence grounding every claim

Amber (Jiachen) Liu 🍊

130,564 views • 3 months ago

🚨I found the Best Blitzes in Madden 25 ⤵️ 🚨 Denver Broncos Defensive Playbook 1⃣ Nickel 245 Odd - Easiest way to scream off the edge sending 3, & the Running back really struggles to block it. Half & Full slide often leaves someone coming right in - Can Get DT A Gap sending 4! - Univeral Blitz Setup that works for every coverage! 2⃣ Dime Sugar Wk - Best Coverage D in Madden 25 while also getting A Gap pressure sending as little as 3 rushers! And thats an A Gap Blitz to a Safety on the line of scrimmage. Talk about screaming 😱 - We use Cover 4 Quarters & Cover 2 Man with these defense because those are the best 2 coverages in Madden 25 right now, especially for regs. 3⃣ Nickel 24 Double Mug - Yes the meta Mid Blitz defense good, but we found something even better! - A Gap sending 3 and you can mix the meta crash down with it with the same exact look All of our Defenses are Found in our Madden Daily Film Room. Link to that in the next post below⤵️

🚨I found the Best Blitzes in Madden 25 ⤵️ 🚨 Denver Broncos Defensive Playbook 1⃣ Nickel 245 Odd - Easiest way to scream off the edge sending 3, & the Running back really struggles to block it. Half & Full slide often leaves someone coming right in - Can Get DT A Gap sending 4! - Univeral Blitz Setup that works for every coverage! 2⃣ Dime Sugar Wk - Best Coverage D in Madden 25 while also getting A Gap pressure sending as little as 3 rushers! And thats an A Gap Blitz to a Safety on the line of scrimmage. Talk about screaming 😱 - We use Cover 4 Quarters & Cover 2 Man with these defense because those are the best 2 coverages in Madden 25 right now, especially for regs. 3⃣ Nickel 24 Double Mug - Yes the meta Mid Blitz defense good, but we found something even better! - A Gap sending 3 and you can mix the meta crash down with it with the same exact look All of our Defenses are Found in our Madden Daily Film Room. Link to that in the next post below⤵️

LitezBreakdowns

32,719 views • 2 years ago

Full story with Armani Ferrante on 24/7 US equities for international investors. What is Backpack? Bona fide security entitlements. Stablecoins. US capital markets. 00:34 What is Backpack? Your neobrokerage. Your money. Your exchange. “Backpack is a globally regulated financial institution for the future of money. And right now, we are in the early innings of this incredible transformation of global finance all around the world.” 01:47 24/7 trading with a bona fide security entitlement 24/7 trading began in crypto. But not all tokenized stocks represent the same asset or come with the same legal rights. “If you take a look at what the tokenized stock issuers are doing today... these properties might be different legal rights. They might be cash settled. And so they might be more akin to something like a CFD rather than an actual security entitlement.” “What we're doing is really for the first time opening up 24/7 trading for a genuine bona fide security entitlement.” “Markets will close on Friday in the US. And over the weekend, Saturday and Sunday, people through the Backpack brokerage will continue to trade and they'll get the exact same asset that you would expect from your neighborhood brokerage account.” 03:47 The world building on US capital markets “Jensen Huang from NVIDIA, he loves to say that the world needs to build on the American tech stack. But we don't talk enough about the world building on the US capital market. And that's really the thing that's happening with crypto right now.” 04:07 Stablecoins first. Tokenized equities next. “Maybe from a domestic point of view, a stablecoin is a story about efficiency, a story about upgrading the existing financial system. But when you look internationally, it's a very different story. It's a story about access. It's a story about the entire world wanting access to US dollars, even if they don't have access to local USD banking rails.” Circle and Tether have grown into major holders of US Treasuries. Backpack is bringing that same global access to tokenized US equities on Solana. “In the exact same way that US dollars have really been exported to the entire world and proliferated through the form factor of a stablecoin, we fully expect tokenization of US equities to have the exact same effect, but for publicly traded companies in the US all around the world.”

Full story with Armani Ferrante on 24/7 US equities for international investors. What is Backpack? Bona fide security entitlements. Stablecoins. US capital markets. 00:34 What is Backpack? Your neobrokerage. Your money. Your exchange. “Backpack is a globally regulated financial institution for the future of money. And right now, we are in the early innings of this incredible transformation of global finance all around the world.” 01:47 24/7 trading with a bona fide security entitlement 24/7 trading began in crypto. But not all tokenized stocks represent the same asset or come with the same legal rights. “If you take a look at what the tokenized stock issuers are doing today... these properties might be different legal rights. They might be cash settled. And so they might be more akin to something like a CFD rather than an actual security entitlement.” “What we're doing is really for the first time opening up 24/7 trading for a genuine bona fide security entitlement.” “Markets will close on Friday in the US. And over the weekend, Saturday and Sunday, people through the Backpack brokerage will continue to trade and they'll get the exact same asset that you would expect from your neighborhood brokerage account.” 03:47 The world building on US capital markets “Jensen Huang from NVIDIA, he loves to say that the world needs to build on the American tech stack. But we don't talk enough about the world building on the US capital market. And that's really the thing that's happening with crypto right now.” 04:07 Stablecoins first. Tokenized equities next. “Maybe from a domestic point of view, a stablecoin is a story about efficiency, a story about upgrading the existing financial system. But when you look internationally, it's a very different story. It's a story about access. It's a story about the entire world wanting access to US dollars, even if they don't have access to local USD banking rails.” Circle and Tether have grown into major holders of US Treasuries. Backpack is bringing that same global access to tokenized US equities on Solana. “In the exact same way that US dollars have really been exported to the entire world and proliferated through the form factor of a stablecoin, we fully expect tokenization of US equities to have the exact same effect, but for publicly traded companies in the US all around the world.”

Backpack 🎒

51,126 views • 1 month ago

THIS BUILDER JUST DROPPED 256GB OF RAM INTO A MONSTER THREADRIPPER WORKSTATION TO RUN UNFILTERED LOCAL AI Imagine trying to fit a computer setup into a chassis that is basically the size of a mini fridge. That is the Corsair 1000D tower. The builder crammed an ASUS Pro WS WRX80E-SAGE motherboard inside and slapped a 64-core AMD Threadripper PRO 5995WX right into the socket Why go this heavy? Simple. To make sure local open weights models don't instantly choke standard desktop hardware during heavy reasoning tasks Then things get downright ridiculous with the memory configuration. He unboxes eight separate Kingston DDR4 modules, pinning a massive 256 gigabytes of system RAM directly to the board. Having that kind of local memory headroom is an absolute necessity if a team wants to handle massive datasets or run dense training loops without constantly swapping data to the storage drives Speaking of storage, the system relies on two lightning fast 2TB Samsung 990 PRO NVMe drives For the graphics pipeline, he drops in a top-tier ASUS ROG RTX 4090 boasting 24 gigabytes of VRAM. That is pretty much the gold standard right now if you want to run quick local inference cycles and completely stop paying corporate cloud token fees to OpenAI or Anthropic Powering this whole grid requires a monstrous ASUS ROG 1600W Thor Gen 2 power supply. And to prevent the entire workstation from turning into a space heater under full load, the builder went all out with a 360mm AIO liquid cooling setup and an insane cluster of sixteen Lian Li SL-Infinity RGB fans It looks incredibly flashy, probably sounds like a jet engine when the cores push maximum load What to buy for local AI? => my guide below Bookmark this so you don't lose it

THIS BUILDER JUST DROPPED 256GB OF RAM INTO A MONSTER THREADRIPPER WORKSTATION TO RUN UNFILTERED LOCAL AI Imagine trying to fit a computer setup into a chassis that is basically the size of a mini fridge. That is the Corsair 1000D tower. The builder crammed an ASUS Pro WS WRX80E-SAGE motherboard inside and slapped a 64-core AMD Threadripper PRO 5995WX right into the socket Why go this heavy? Simple. To make sure local open weights models don't instantly choke standard desktop hardware during heavy reasoning tasks Then things get downright ridiculous with the memory configuration. He unboxes eight separate Kingston DDR4 modules, pinning a massive 256 gigabytes of system RAM directly to the board. Having that kind of local memory headroom is an absolute necessity if a team wants to handle massive datasets or run dense training loops without constantly swapping data to the storage drives Speaking of storage, the system relies on two lightning fast 2TB Samsung 990 PRO NVMe drives For the graphics pipeline, he drops in a top-tier ASUS ROG RTX 4090 boasting 24 gigabytes of VRAM. That is pretty much the gold standard right now if you want to run quick local inference cycles and completely stop paying corporate cloud token fees to OpenAI or Anthropic Powering this whole grid requires a monstrous ASUS ROG 1600W Thor Gen 2 power supply. And to prevent the entire workstation from turning into a space heater under full load, the builder went all out with a 360mm AIO liquid cooling setup and an insane cluster of sixteen Lian Li SL-Infinity RGB fans It looks incredibly flashy, probably sounds like a jet engine when the cores push maximum load What to buy for local AI? => my guide below Bookmark this so you don't lose it

beamnxw ./

47,366 views • 1 month ago

Dr. Amy Shah just revealed one of the simplest, most powerful brain hacks: Walking just 40 minutes, three times a week can literally grow your hippocampus — the area of your brain tied to memory, learning, and mood — even as an adult. Multiple neuroimaging studies (including MRI research) show that regular aerobic walking increases hippocampal volume by 1–2% in adults, which correlates with better memory performance and reduced risk of cognitive decline. She especially recommends it for women in midlife because it lowers cortisol, improves mood, and gives your nervous system the recovery it desperately needs. In a world full of expensive supplements and complex protocols, something as free and basic as walking delivers real, measurable changes to your brain. This one hit me hard. I’ve always known walking helps my mood, but hearing it can actually grow brain tissue makes me way more consistent about protecting that time — especially outdoors. Do you have a regular walking habit, or does this make you want to start one?

Dr. Amy Shah just revealed one of the simplest, most powerful brain hacks: Walking just 40 minutes, three times a week can literally grow your hippocampus — the area of your brain tied to memory, learning, and mood — even as an adult. Multiple neuroimaging studies (including MRI research) show that regular aerobic walking increases hippocampal volume by 1–2% in adults, which correlates with better memory performance and reduced risk of cognitive decline. She especially recommends it for women in midlife because it lowers cortisol, improves mood, and gives your nervous system the recovery it desperately needs. In a world full of expensive supplements and complex protocols, something as free and basic as walking delivers real, measurable changes to your brain. This one hit me hard. I’ve always known walking helps my mood, but hearing it can actually grow brain tissue makes me way more consistent about protecting that time — especially outdoors. Do you have a regular walking habit, or does this make you want to start one?

Camus

24,095 views • 3 months ago

🔴 INTRODUCING BLOOD v3 🔴 The most advanced Solana bot just got even better. After months of relentless work, testing, and fine-tuning, we're proud to unveil BLOOD GUI – the most intuitive, lightning-fast, and powerful UI ever created. Built for speed, simplicity, and precision, it redefines the way you interact with bot. Newcomers can set everything up in minutes, while pros will execute their strategies with unmatched efficiency. 🔥 WHAT’S NEW? 🔥 ✅ Informative homepage – get all essential bot details at a glance. ✅ Sleek, optimized UI – clean design with zero performance loss. ✅ Smart settings management – hide unnecessary options or find the right setting instantly with a search bar. ✅ Enhanced position control – manage multiple trading modes with full control over your trades. ✅ 🔁 Unlimited wallet copying – mirror an infinite number of wallets. ✅ ⚡️ Upgraded nodes – FREE NODE + FASTEST GEYSERS FOR ALL HOLDERS. ✅ 📊 PNL & Winrate analytics – track performance, optimize strategies, and maximize profits with in-depth wallet analysis. And the best part? You get ALL of this with ZERO TRADING TAXES 💰. 🚨 This is not just an update. It’s a GAME-CHANGER. 🚨 There is also one more important thing…

🔴 INTRODUCING BLOOD v3 🔴 The most advanced Solana bot just got even better. After months of relentless work, testing, and fine-tuning, we're proud to unveil BLOOD GUI – the most intuitive, lightning-fast, and powerful UI ever created. Built for speed, simplicity, and precision, it redefines the way you interact with bot. Newcomers can set everything up in minutes, while pros will execute their strategies with unmatched efficiency. 🔥 WHAT’S NEW? 🔥 ✅ Informative homepage – get all essential bot details at a glance. ✅ Sleek, optimized UI – clean design with zero performance loss. ✅ Smart settings management – hide unnecessary options or find the right setting instantly with a search bar. ✅ Enhanced position control – manage multiple trading modes with full control over your trades. ✅ 🔁 Unlimited wallet copying – mirror an infinite number of wallets. ✅ ⚡️ Upgraded nodes – FREE NODE + FASTEST GEYSERS FOR ALL HOLDERS. ✅ 📊 PNL & Winrate analytics – track performance, optimize strategies, and maximize profits with in-depth wallet analysis. And the best part? You get ALL of this with ZERO TRADING TAXES 💰. 🚨 This is not just an update. It’s a GAME-CHANGER. 🚨 There is also one more important thing…

BLOOD SOLUTIONS

12,258 views • 1 year ago

🔥HOLY SMOKES! $TAO holders! 🚀 SUBNET 19 (VISION) ON BITTENSOR IS ABSOLUTELY CRUSHING IT! In my 5+ years covering crypto and AI, this is one of the most impressive implementations I've seen. The combination of scale, performance, and decentralization is absolutely next level! 🚀 @namoray_dev @Corcel_X 💨 INSANE Speed Performance: - Llama 3.1 8B: 196.18 tokens/s with +107.23% advantage - Llama 3.1 70B: 124.96 tokens/s with +154.96% advantage - Llama 3.2 3B: 166.69 tokens/s with +21.66% advantage 🔥 Top Tier Model Integration: - Meta-Llama-3-70B & 8B Instruct - FLUX.1-schnell for Text-to-Image - ProteusV0.4-Lightning (Text & Image) - Multiple model variations for redundancy 🔥 What Makes This INSANE: - Complete decentralization - No single point of failure - Multiple model choices for redundancy - Real-time performance tracking - Transparent incentive structure The incentive distribution curve shows a healthy network with: - Strong rewards for top performers - Fair distribution across all participants - Clear path for growth and improvement - Sustainable economic model What's truly MIND-BLOWING is how they've managed to: 1. Scale to millions of operations 2. Maintain high quality across multiple tasks 3. Create a fair, competitive marketplace 4. Build in redundancy and reliability 5. Achieve true decentralization This isn't just another subnet - this is the future of decentralized AI inference happening RIGHT NOW! 🔥 1. MASSIVE Scale & Adoption: - We're seeing 7M+ tokens being processed - 14K+ processing steps being executed - Multiple AI models running simultaneously - Incredible miner participation across the network 2. Revolutionary Task Distribution: - Llama 3.1 70B leading with 20% weighting - Avatar Generation at 15% - Perfectly balanced task distribution for optimal network performance - Multiple specialized tasks including Text-to-Image and Image-to-Image processing 3. Elite Performance Metrics: - Top miners hitting 0.00775 incentive rates - Consistent performance across the network - Impressive scaling from top to bottom performers - Strong incentive curve maintaining network quality 📈 Network Performance: - Consistent upward trend in tokens/s - Quality scores maintaining high levels (>0.9) - Steady improvement in miner performance - Rock-solid network reliability ⚡ Platform Highlights: - Permissionless, serverless architecture - Global network of Always-On GPUs - Instant API access - Full decentralization - Multi-model support with seamless switching What makes this TRULY SPECIAL is the consistent upward trajectory in both speed and quality, while maintaining a decentralized architecture. The performance advantages over industry standards (+154.96% for 70B!) are absolutely mind-blowing! 🚀 This isn't just another AI subnet - it's a glimpse into the future of decentralized AI inference! The combination of speed, reliability, and model variety makes this one of the most impressive implementations in the space! 🔥 📽 Watch Now on YouTube and TikTok: Source 🔗

Andy ττ

11,616 views • 1 year ago

I'm using macOS for more than 14 years now, in a single laptop screen setup with lots of windows open, and I'm using macOS spaces everyday to organize my windows. Switching between spaces is easy: I hit CTRL+arrow left/right, it takes no time. What takes time however, is this macOS-native transition animation that takes a bit less than 1 second, for every single space switch. I've checked everywhere in macOS settings, asked friends with the same problem, and there's just no setting to disable this effect (even with Reduced Motion ON in Accessibility settings). This UX choice from Apple alone costs me days of my time every year, as I'm working non-stop on my SaaS business for 10h a day 5-6 days a week (Crisp for those who know about it!). I calculated that with my daily use of macOS spaces, in 6 months I'd perform 100,000 switches, which costs me the same in seconds waiting for the transition to complete. This represents almost 28 hours waiting. In 1 year, I'd have lost 2.4 days just... waiting! The Hack Fortunately, some other open-source projects discovered a hack not so long ago, which allows disabling the transition when switching spaces: it involves sending to the system a "fake" trackpad left/right swipe gesture that's faster than what's expected (faster than what a human can input via the Macbook's trackpad). Above some threshold, macOS just perform the space switch with absolutely no transition. A friend of mine, Yaël 🛸, made a quick prototype in Objective-C, and it magically worked! The Solution So, I asked my friend if we could make it a proper macOS app together, in a beautiful minimalist package, and he said yes! So, we launched Space Rabbit, a tiny 2MB fully native macOS app, that allows you to optimize macOS spaces for productivity! Space Rabbit is a full rewrite of the initial prototype in modern Swift, with a tray menu bar to configure the app, and a settings window, in a macOS app packaging. Space Rabbit can be downloaded right here: We just released V2 of the app last week, fixing a ton of little bugs, especially for dual screen setups. It's actively maintained to this day as it's my daily driver. The app is open-source: and will always be free, as it's helping improve our productivity in our daily Web Dev tasks for our main businesses. We'll be happy to implement more features from user requests, if you find this app useful and use it daily!

I'm using macOS for more than 14 years now, in a single laptop screen setup with lots of windows open, and I'm using macOS spaces everyday to organize my windows. Switching between spaces is easy: I hit CTRL+arrow left/right, it takes no time. What takes time however, is this macOS-native transition animation that takes a bit less than 1 second, for every single space switch. I've checked everywhere in macOS settings, asked friends with the same problem, and there's just no setting to disable this effect (even with Reduced Motion ON in Accessibility settings). This UX choice from Apple alone costs me days of my time every year, as I'm working non-stop on my SaaS business for 10h a day 5-6 days a week (Crisp for those who know about it!). I calculated that with my daily use of macOS spaces, in 6 months I'd perform 100,000 switches, which costs me the same in seconds waiting for the transition to complete. This represents almost 28 hours waiting. In 1 year, I'd have lost 2.4 days just... waiting! The Hack Fortunately, some other open-source projects discovered a hack not so long ago, which allows disabling the transition when switching spaces: it involves sending to the system a "fake" trackpad left/right swipe gesture that's faster than what's expected (faster than what a human can input via the Macbook's trackpad). Above some threshold, macOS just perform the space switch with absolutely no transition. A friend of mine, Yaël 🛸, made a quick prototype in Objective-C, and it magically worked! The Solution So, I asked my friend if we could make it a proper macOS app together, in a beautiful minimalist package, and he said yes! So, we launched Space Rabbit, a tiny 2MB fully native macOS app, that allows you to optimize macOS spaces for productivity! Space Rabbit is a full rewrite of the initial prototype in modern Swift, with a tray menu bar to configure the app, and a settings window, in a macOS app packaging. Space Rabbit can be downloaded right here: We just released V2 of the app last week, fixing a ton of little bugs, especially for dual screen setups. It's actively maintained to this day as it's my daily driver. The app is open-source: and will always be free, as it's helping improve our productivity in our daily Web Dev tasks for our main businesses. We'll be happy to implement more features from user requests, if you find this app useful and use it daily!

Valerian Saliou

219,111 views • 20 days ago

single RTX 3090. 24 GB VRAM. Qwen3.5-35B-A3B. 4-bit quant, 113 tokens per second at full 262K context harnessing Claude Code locally with no API, no subscription, no proxy. told it what it is. 30 Mamba2 layers, 10 attention, 256 experts, 8 active per token. said "build something that shows off what you can do." it visualized its own architecture. interactive. tokens flowing through layers. 256 experts lighting up on routing. served in the browser from the same GPU running inference. single prompt. then i said level up. 3D. Three.js. separate files. flythrough camera. clickable layers. it planned first, scaffolded 6 files, hit one API bug, fixed it itself, then optimized for smooth framerate. two iterations to a working 3D neural network explorer. llama.cpp just merged a native Anthropic endpoint. Claude Code points at localhost. the whole setup is two commands. no LiteLLM. no proxy config. the open source models coming out of china right now are genuinely changing what's possible on consumer hardware. respect to the Qwen team. this is acceleration.

single RTX 3090. 24 GB VRAM. Qwen3.5-35B-A3B. 4-bit quant, 113 tokens per second at full 262K context harnessing Claude Code locally with no API, no subscription, no proxy. told it what it is. 30 Mamba2 layers, 10 attention, 256 experts, 8 active per token. said "build something that shows off what you can do." it visualized its own architecture. interactive. tokens flowing through layers. 256 experts lighting up on routing. served in the browser from the same GPU running inference. single prompt. then i said level up. 3D. Three.js. separate files. flythrough camera. clickable layers. it planned first, scaffolded 6 files, hit one API bug, fixed it itself, then optimized for smooth framerate. two iterations to a working 3D neural network explorer. llama.cpp just merged a native Anthropic endpoint. Claude Code points at localhost. the whole setup is two commands. no LiteLLM. no proxy config. the open source models coming out of china right now are genuinely changing what's possible on consumer hardware. respect to the Qwen team. this is acceleration.

Sudo su

110,206 views • 6 months ago

REPORT: Big Tech is moving into the one thing everyone depends on: the food supply—and it’s happening right now, in plain sight. What’s being sold as innovation is becoming the foundation of a system that’s far more centralized, and far harder to walk away from. Peter Thiel just poured $2 billion into a company called Halter, where farmers can control entire herds through an app. Tap a button, and hundreds of thousands of cows move on command, guided by solar-powered collars tracking their health, behavior, and location 24/7 with machine learning. Now layer in the 2026 Farm Bill, which covers up to 90% of the cost for adopting these systems, while letting tech companies, not regulators, define how they operate. Food, data, infrastructure… all being pulled into the same network. At that point, the question isn’t whether this makes farming easier. It’s who controls the system once everything runs through it, and what happens when access to it depends on compliance. And once you see how this actually works, it’s hard to look at any of this the same way again. Watch ZeeeMediaOfficial's full report: 👇

REPORT: Big Tech is moving into the one thing everyone depends on: the food supply—and it’s happening right now, in plain sight. What’s being sold as innovation is becoming the foundation of a system that’s far more centralized, and far harder to walk away from. Peter Thiel just poured $2 billion into a company called Halter, where farmers can control entire herds through an app. Tap a button, and hundreds of thousands of cows move on command, guided by solar-powered collars tracking their health, behavior, and location 24/7 with machine learning. Now layer in the 2026 Farm Bill, which covers up to 90% of the cost for adopting these systems, while letting tech companies, not regulators, define how they operate. Food, data, infrastructure… all being pulled into the same network. At that point, the question isn’t whether this makes farming easier. It’s who controls the system once everything runs through it, and what happens when access to it depends on compliance. And once you see how this actually works, it’s hard to look at any of this the same way again. Watch ZeeeMediaOfficial's full report: 👇

The Vigilant Fox 🦊

57,127 views • 5 months ago

🚨 BLACK FRIDAY ANNOUNCEMENT 🚨 YieldBOSS is OPEN after $23M in TOTAL SERVER TRADING GAINS since inception — and we’re celebrating with our BIGGEST DEAL EVER. 🎉 💵 $100 OFF your first month 🏷️ Type Promo Code: YIELDBOSSBF ⏳ ONLY 75 SPOTS OPEN 👨‍👩‍👧 Pro, family-centric community of hundreds of consistently profitable traders 📚 Full step-by-step educational course (ranging from the basics -> advanced proprietary methodologies) 🕰️ Access to the POI System 🎯 Access to the Target Plane System in its' entirety ☁️ Access to our proprietary Cloud System indicator (free w/ membership) 📈 Daily insights / analysis / picks 📊 Premium Options Data 24/7 (Flow, GEX, and more) Our members earn while they learn and report seeing consistent profitability within 3-6 months through the rigorous education, data, and analysis we provide, — including an entire step-by-step course of the Target Plane System and countless other strategies / resources. Due to our extremely low member-churn rate, we know when people come; they stay. This is our way of giving back to the retail community and trying to help as many aspiring traders and investors as possible. Want to finally trade with an edge & be surrounded by winners? Join our family of 450+ members and use promo code 🔑 "YIELDBOSSBF" for $100 OFF your first month. DM Nicholas A. Pantano or Darren (YieldBOSS) to apply for Membership; let's get to work. ⚡

🚨 BLACK FRIDAY ANNOUNCEMENT 🚨 YieldBOSS is OPEN after $23M in TOTAL SERVER TRADING GAINS since inception — and we’re celebrating with our BIGGEST DEAL EVER. 🎉 💵 $100 OFF your first month 🏷️ Type Promo Code: YIELDBOSSBF ⏳ ONLY 75 SPOTS OPEN 👨‍👩‍👧 Pro, family-centric community of hundreds of consistently profitable traders 📚 Full step-by-step educational course (ranging from the basics -> advanced proprietary methodologies) 🕰️ Access to the POI System 🎯 Access to the Target Plane System in its' entirety ☁️ Access to our proprietary Cloud System indicator (free w/ membership) 📈 Daily insights / analysis / picks 📊 Premium Options Data 24/7 (Flow, GEX, and more) Our members earn while they learn and report seeing consistent profitability within 3-6 months through the rigorous education, data, and analysis we provide, — including an entire step-by-step course of the Target Plane System and countless other strategies / resources. Due to our extremely low member-churn rate, we know when people come; they stay. This is our way of giving back to the retail community and trying to help as many aspiring traders and investors as possible. Want to finally trade with an edge & be surrounded by winners? Join our family of 450+ members and use promo code 🔑 "YIELDBOSSBF" for $100 OFF your first month. DM Nicholas A. Pantano or Darren (YieldBOSS) to apply for Membership; let's get to work. ⚡

Darren (YieldBOSS)

18,511 views • 9 months ago

Hermes agent just left the terminal. 𝗛𝗲𝗿𝗺𝗲𝘀 𝗗𝗲𝘀𝗸𝘁𝗼𝗽 dropped yesterday. native app for macOS, Windows, and Linux. for months Hermes was the agent that learned your projects, wrote its own skills, and built a model of who you are. all of it buried in terminal logs. now it has a window. the important part is that it's not a wrapper. it runs the same agent core, the same sessions, memory, and skills as the CLI. you can start a task in the terminal and finish it in the app without anything resetting. the state is shared across every interface, not copied between them. what the GUI actually adds: → streaming chat that shows live tool calls and inline reasoning instead of a spinner → a preview rail that renders pages, code, and images right beside the conversation → an artifacts panel that collects every file the agent has ever produced → remote gateway mode, so you can point the app at a VPS and run the heavy work elsewhere → skills, cron, profiles, and gateways managed point-and-click instead of through YAML → voice mode, drag-drop files, and inline image generation remote gateway mode is the one worth slowing down on. the agent runs 24/7 on a $5 server while you control it from your laptop like a local app. other agent UIs are chatboxes with a logo. this one shows the autonomy instead of hiding it, so you watch the skills load, the tools fire, and the artifacts pile up as it works. it was teased in Jensen's GTC keynote. MIT licensed, local-first, no telemetry. if you already run Hermes, download it and everything is already there. your chats, memory, and skills carry straight over. i wrote a full masterclass on Hermes Agent that walks through the SOUL. md identity layer, the three-tier memory system, the self-evolving skills loop, and how to run three specialized agents 24/7. desktop is the interface that finally does all of it justice. the article is quoted below.

Akshay 🚀

51,540 views • 2 months ago

🚨🇦🇺🇨🇳 INTERVIEW: EVER WONDERED WHAT IT’S LIKE BEING ARRESTED IN CHINA? This is one of the wildest stories I’ve ever heard Tl;dr: It’s terrifying, and should make anyone in the West grateful and appreciative of their legal system Australian citizen Matthew Radalj was arrested for stealing his own phone… yep, you heard that right, he stole his own phone 😂 He was then beaten, tortured, and jailed for years And his embassy? Barely helped. They apparently focus on helping Australians born in China more than Australians born in Australia Mathew’s story: He breaks his iPhone 11 and takes it to a local shop he trusted, with a guy he’d been going to for years. The guys tries to charge him more than than he initially quoted Mathew refuses, and pays the man what they initially agreed, and walks off Next thing he knows? The security pepper sprays him, he fights back, people mob him…. And he gets arrested After talking to police and going to the station, he finds out they didn’t care for the facts and already consider him the ‘foreign villain’, and he somehow ends up getting beaten, then tortured and jailed for years. It’s one of those stories that makes you rethink how safe you really are when you’re overseas, even if you’ve done “everything right.” And if you think your embassy will always have your back abroad… think again. Full Interview with @MatthewRadalj below. 0:00 - How taking his own phone led to 5 years in a Chinese prison 3:12 - A normal phone repair that spiraled fast 7:05 - Getting jumped by security and chased by a mob in Beijing 8:53 - Police show up and suddenly money becomes the focus 11:28 - Finding out robbery is the official story 13:03 - Getting pulled into a death accusation out of nowhere 14:06 - When being Australian suddenly matters 17:23 - What 2 days of custody really looked like 18:05 - Being stripped and processed into the detention system 21:21 - Seeing people strapped to chairs and trying to make sense of it 22:33 - Being questioned while badly concussed 24:27 - The moment it sinks in that this is prison 25:24 - Walking into a Chinese prison cell 26:00 - The simple question no one ever asked about the phone 27:31 - How the story locks in early 30:09 - Asking for a lawyer and hitting a wall 31:05 - The Australian Embassy finally visits 31:24 - That visit going in a very different direction 33:28 - What 5 years inside does to your trust 34:18 - Realizing the system was the only link out

🚨🇦🇺🇨🇳 INTERVIEW: EVER WONDERED WHAT IT’S LIKE BEING ARRESTED IN CHINA? This is one of the wildest stories I’ve ever heard Tl;dr: It’s terrifying, and should make anyone in the West grateful and appreciative of their legal system Australian citizen Matthew Radalj was arrested for stealing his own phone… yep, you heard that right, he stole his own phone 😂 He was then beaten, tortured, and jailed for years And his embassy? Barely helped. They apparently focus on helping Australians born in China more than Australians born in Australia Mathew’s story: He breaks his iPhone 11 and takes it to a local shop he trusted, with a guy he’d been going to for years. The guys tries to charge him more than than he initially quoted Mathew refuses, and pays the man what they initially agreed, and walks off Next thing he knows? The security pepper sprays him, he fights back, people mob him…. And he gets arrested After talking to police and going to the station, he finds out they didn’t care for the facts and already consider him the ‘foreign villain’, and he somehow ends up getting beaten, then tortured and jailed for years. It’s one of those stories that makes you rethink how safe you really are when you’re overseas, even if you’ve done “everything right.” And if you think your embassy will always have your back abroad… think again. Full Interview with @MatthewRadalj below. 0:00 - How taking his own phone led to 5 years in a Chinese prison 3:12 - A normal phone repair that spiraled fast 7:05 - Getting jumped by security and chased by a mob in Beijing 8:53 - Police show up and suddenly money becomes the focus 11:28 - Finding out robbery is the official story 13:03 - Getting pulled into a death accusation out of nowhere 14:06 - When being Australian suddenly matters 17:23 - What 2 days of custody really looked like 18:05 - Being stripped and processed into the detention system 21:21 - Seeing people strapped to chairs and trying to make sense of it 22:33 - Being questioned while badly concussed 24:27 - The moment it sinks in that this is prison 25:24 - Walking into a Chinese prison cell 26:00 - The simple question no one ever asked about the phone 27:31 - How the story locks in early 30:09 - Asking for a lawyer and hitting a wall 31:05 - The Australian Embassy finally visits 31:24 - That visit going in a very different direction 33:28 - What 5 years inside does to your trust 34:18 - Realizing the system was the only link out

Mario Nawfal

979,696 views • 6 months ago

$I just crammed the updated Gemma 4 26B A4B QAT (MoE) with 180k context into an 8GB RTX 4060 (8 GB VRAM + 16 GB RAM only!!) and optimized the batch size. 23 tokens/sec decode, 300 tokens/sec prefill Yesterday I showed you a Gemma 4 31B dense model running flawlessly on an RTX 4090. Today, we're breaking the VRAM bank on a budget card using Unsloth’s new Gemma 4 26B (A4B) QAT quants. Following Google’s chat template update that boosted agentic benchmarks by +10%, I pushed this model to its absolute limits. Here is how you squeeze 250k context out of 8GB of VRAM. # The Setup & The Optimization - Hardware: Nvidia RTX 4060 (8GB VRAM) + 16GB System RAM - Environment: CUDA 13.0 build of llama.cpp - Model: gemma-4-26B-A4B-it-qat-UD-Q4_K_XL.gguf - Prompt: 28,000 tokens of prompt for each run If you read my L2 cache breakdown (attached in replies), you know the 4060’s 24MB cache maxes out at `-b 1024 -ub 1024`. Push past that, and prefill crashes. I locked those flags in for every test below to ensure maximum GEMM throughput. # 1. The Raw Context Push (Unquantized KV Cache) First, I wanted to see how far pure 8GB VRAM + 16GB RAM could stretch without touching the KV cache: - 80k Context: Prefill 385 t/s | Decode 25.5 t/s - 120k Context: Prefill 270 t/s | Decode 24 t/s llama.cpp flags: .\llama-server -m gemma-4-26B-A4B-it-qat-UD-Q4_K_XL.gguf -c 120000 --port 8080 -ub 1024 -b 1024 Without KV quantization, 120k is your hard ceiling. push past that prefill throughput drops off a cliff, making the model practically unusable for large agentic workloads. # 2. The Q8 KV Cache Lifeline To survive 250k context on a budget card, you have to quantize the KV cache. I enabled 8 bit KV cache (`-ctk q8_0 -ctv q8_0`) and re ran: - 180k Context: Prefill 280 t/s | Decode 22.8 t/s - 250k Context: Prefill 115 t/s | Decode 20 t/s llama.cpp flags: .\llama-server -m gemma-4-26B-A4B-it-qat-UD-Q4_K_XL.gguf -c 180000 --port 8080 -b 1024 -ub 1024 -ctk q8_0 -ctv q8_0 Result: Q8 KV cache brings 250k context back from the dead. Decode speed stabilizes at a highly usable 20 t/s. You are trading a very small bit amount of reasoning precision for an extra 130,000 tokens of context window. if you own a single rtx 3050, 3060, 3070, 4050, 4060, 5050 or 5060, you must try this model and optimize your batch size for higher prefill. Hugging Face links to the updated Unsloth's QAT quants and performance graph are in the replies below. What model are you running on your 6GB, 8GB or 12GB cards right now? Let's see your setups.$

I just crammed the updated Gemma 4 26B A4B QAT (MoE) with 180k context into an 8GB RTX 4060 (8 GB VRAM + 16 GB RAM only!!) and optimized the batch size. 23 tokens/sec decode, 300 tokens/sec prefill Yesterday I showed you a Gemma 4 31B dense model running flawlessly on an RTX 4090. Today, we're breaking the VRAM bank on a budget card using Unsloth’s new Gemma 4 26B (A4B) QAT quants. Following Google’s chat template update that boosted agentic benchmarks by +10%, I pushed this model to its absolute limits. Here is how you squeeze 250k context out of 8GB of VRAM. # The Setup & The Optimization - Hardware: Nvidia RTX 4060 (8GB VRAM) + 16GB System RAM - Environment: CUDA 13.0 build of llama.cpp - Model: gemma-4-26B-A4B-it-qat-UD-Q4_K_XL.gguf - Prompt: 28,000 tokens of prompt for each run If you read my L2 cache breakdown (attached in replies), you know the 4060’s 24MB cache maxes out at `-b 1024 -ub 1024`. Push past that, and prefill crashes. I locked those flags in for every test below to ensure maximum GEMM throughput. # 1. The Raw Context Push (Unquantized KV Cache) First, I wanted to see how far pure 8GB VRAM + 16GB RAM could stretch without touching the KV cache: - 80k Context: Prefill 385 t/s | Decode 25.5 t/s - 120k Context: Prefill 270 t/s | Decode 24 t/s llama.cpp flags: .\llama-server -m gemma-4-26B-A4B-it-qat-UD-Q4_K_XL.gguf -c 120000 --port 8080 -ub 1024 -b 1024 Without KV quantization, 120k is your hard ceiling. push past that prefill throughput drops off a cliff, making the model practically unusable for large agentic workloads. # 2. The Q8 KV Cache Lifeline To survive 250k context on a budget card, you have to quantize the KV cache. I enabled 8 bit KV cache (`-ctk q8_0 -ctv q8_0`) and re ran: - 180k Context: Prefill 280 t/s | Decode 22.8 t/s - 250k Context: Prefill 115 t/s | Decode 20 t/s llama.cpp flags: .\llama-server -m gemma-4-26B-A4B-it-qat-UD-Q4_K_XL.gguf -c 180000 --port 8080 -b 1024 -ub 1024 -ctk q8_0 -ctv q8_0 Result: Q8 KV cache brings 250k context back from the dead. Decode speed stabilizes at a highly usable 20 t/s. You are trading a very small bit amount of reasoning precision for an extra 130,000 tokens of context window. if you own a single rtx 3050, 3060, 3070, 4050, 4060, 5050 or 5060, you must try this model and optimize your batch size for higher prefill. Hugging Face links to the updated Unsloth's QAT quants and performance graph are in the replies below. What model are you running on your 6GB, 8GB or 12GB cards right now? Let's see your setups.

Alok

36,617 views • 1 month ago

AMD CEO Lisa Su just killed Nvidia’s $4,000 AI box with a $1,499 lunchbox. She walked on stage, held it in one hand, and ran a 235 billion parameter model live. No data center. No cloud. No rented GPU. The chip inside is something nobody saw coming. AMD’s Ryzen AI Max+ 395 is the first x86 silicon where CPU and GPU share the same 128GB of memory. That single trick lets a desktop run models that used to need a server rack. Out of those 128GB, Linux hands the GPU 110GB to play with. For context, an RTX 5090 gives you 32GB. A 4090 gives you 24. This box gives you more than three times either of them, in a chassis the size of a thick paperback. The benchmark that broke the room: this chip beat an Nvidia RTX 5080 by more than 3x on DeepSeek R1 inference. A $1,499 lunchbox outrunning a $1,000 discrete graphics card on a real AI workload. Nvidia spent a decade convincing the world you needed their hardware for serious AI. AMD just put that on a desk for half the price. Here is what nobody is telling you. A heavy AI user right now pays $200 for Claude Code Max, $200 for ChatGPT Pro, $20 for Cursor, $20 for Gemini. That is $5,280 a year leaving your account. The box pays itself off in 9 months and then runs free for the rest of its life. Install Ollama. Pull Qwen3 235B. Point Claude Code at localhost. Same interface you already use, except now nothing leaves your machine, nothing costs per request, and no company throttles your usage at 3am when you finally have time to build. This is the moment every AI subscription becomes optional. Lawyers stop fearing OpenAI leaks. Developers stop watching the token meter. Founders stop renting H100s for prototypes that never ship because the bill scared them. The first thousand people to figure this out will own the next two years of private AI consulting. Save this, and read the full breakdown article below you are watching the next shift hit before everyone else does.

AMD CEO Lisa Su just killed Nvidia’s $4,000 AI box with a $1,499 lunchbox. She walked on stage, held it in one hand, and ran a 235 billion parameter model live. No data center. No cloud. No rented GPU. The chip inside is something nobody saw coming. AMD’s Ryzen AI Max+ 395 is the first x86 silicon where CPU and GPU share the same 128GB of memory. That single trick lets a desktop run models that used to need a server rack. Out of those 128GB, Linux hands the GPU 110GB to play with. For context, an RTX 5090 gives you 32GB. A 4090 gives you 24. This box gives you more than three times either of them, in a chassis the size of a thick paperback. The benchmark that broke the room: this chip beat an Nvidia RTX 5080 by more than 3x on DeepSeek R1 inference. A $1,499 lunchbox outrunning a $1,000 discrete graphics card on a real AI workload. Nvidia spent a decade convincing the world you needed their hardware for serious AI. AMD just put that on a desk for half the price. Here is what nobody is telling you. A heavy AI user right now pays $200 for Claude Code Max, $200 for ChatGPT Pro, $20 for Cursor, $20 for Gemini. That is $5,280 a year leaving your account. The box pays itself off in 9 months and then runs free for the rest of its life. Install Ollama. Pull Qwen3 235B. Point Claude Code at localhost. Same interface you already use, except now nothing leaves your machine, nothing costs per request, and no company throttles your usage at 3am when you finally have time to build. This is the moment every AI subscription becomes optional. Lawyers stop fearing OpenAI leaks. Developers stop watching the token meter. Founders stop renting H100s for prototypes that never ship because the bill scared them. The first thousand people to figure this out will own the next two years of private AI consulting. Save this, and read the full breakdown article below you are watching the next shift hit before everyone else does.

AdiiX

3,394,028 views • 2 months ago

I freaked out when my WiFi router suddenly died. then realized my autonomous Hermes agent is running fully local, nothing stopped. Hermes Agent + Gemma 4 26B A4B QAT MoE, 100% local on my laptop, building my side projects while I scroll my phone zero API calls. zero cost. 100% private. fully offline. This might be the most satisfying thing I’ve watched in a while. last post: showed Hermes + local Gemma 4 26B pull off backtest a trading strategy. this time I asked it to develop something i'd use myself everyday: # A full unpacked extension with: - React side panel UI - Local llama.cpp backend (offline AI) - Live tab sync + status tracking - Auto context extraction via Readability.js Vision on Demand → captures viewport screenshots as compressed JPEGs Deterministic action system -> model outputs tokens -> directly controls page scrolling It planned everything first. Then started executing step by step. all i did was say 'ok'. only once. # What’s wild: - It reports back after every phase - Auto compresses context when nearing limits - Actualy, stays on track llama.cpp flags: -m gemma-4-26B-A4B-it-qat-UD-Q4_K_XL.gguf -c 64000 --cache-type-k q8_0 --cache-type-v q8_0 --port 8080 # Performance on a single NVIDIA RTX 4060 (8GB VRAM) + 16 GB DDR4 RAM Gaming Laptop: - 300 tokens/sec prefill - 25+ tokens/sec decode More than usable for real dev workflows. This isn’t AI demo territory anymore. This is autonomous local software actually building things.

I freaked out when my WiFi router suddenly died. then realized my autonomous Hermes agent is running fully local, nothing stopped. Hermes Agent + Gemma 4 26B A4B QAT MoE, 100% local on my laptop, building my side projects while I scroll my phone zero API calls. zero cost. 100% private. fully offline. This might be the most satisfying thing I’ve watched in a while. last post: showed Hermes + local Gemma 4 26B pull off backtest a trading strategy. this time I asked it to develop something i'd use myself everyday: # A full unpacked extension with: - React side panel UI - Local llama.cpp backend (offline AI) - Live tab sync + status tracking - Auto context extraction via Readability.js Vision on Demand → captures viewport screenshots as compressed JPEGs Deterministic action system -> model outputs tokens -> directly controls page scrolling It planned everything first. Then started executing step by step. all i did was say 'ok'. only once. # What’s wild: - It reports back after every phase - Auto compresses context when nearing limits - Actualy, stays on track llama.cpp flags: -m gemma-4-26B-A4B-it-qat-UD-Q4_K_XL.gguf -c 64000 --cache-type-k q8_0 --cache-type-v q8_0 --port 8080 # Performance on a single NVIDIA RTX 4060 (8GB VRAM) + 16 GB DDR4 RAM Gaming Laptop: - 300 tokens/sec prefill - 25+ tokens/sec decode More than usable for real dev workflows. This isn’t AI demo territory anymore. This is autonomous local software actually building things.

Alok

56,428 views • 1 month ago

BREAKING: $2.5 trillion just got wiped from global markets in 24 hours. Bitcoin crashed below $60,000. The Nasdaq dropped 4% in a single session. And every retail trader's feed is full of people posting their losses in real time... Bitcoin fell from $78,000 to $59,300 in 5 days. A 24% drop in under a week. $1.5 billion in leveraged crypto positions got liquidated in 24 hours. Long positions made up $1.28 billion of that. The fourth time in 5 days that daily liquidations crossed $1 billion. US spot Bitcoin ETFs have now posted 11 consecutive days of outflows. $3.5 billion pulled out by institutions in less than two weeks. Then the equity side cracked. Broadcom dropped 13%. Micron dropped 9.5%. SanDisk dropped 11%. Western Digital dropped 8.5%. Roughly $1 trillion in market cap erased from semiconductors alone. Strategy, the largest corporate Bitcoin holder on earth, dropped 10% after disclosing its first Bitcoin sale in years. 32 coins. $2.5 million. A rounding error against their $59 billion stack. It didn't matter. When the maximalist sells one coin, everybody else hears permission to dump everything. Now here's what actually triggered all of this. The May jobs report came in too strong. 172,000 new jobs against an 85,000 estimate. Prior months revised up by another 93,000. That single data point killed the rate cut narrative. If the economy is too strong, the Fed doesn't cut. If the Fed doesn't cut, risk assets get repriced. Crypto first. Tech stocks second. Everything else last. The entire rally of the past 12 months was built on one assumption. Rates are coming down. That assumption got vaporized in one report. And the market reacted exactly the way it always reacts: Panic. Right now Twitter is full of people posting their losses. "Down $40,000 this week." "Sold my entire crypto position." "Out of the market until things calm down." This is the exact moment retail always loses. Not because the market crashed. Because they're making decisions while watching the crash happen in real time. Every single one of these crashes follows the same pattern. October 2025: $19 billion in liquidations. The biggest wipeout in crypto history. Bitcoin made a new all-time high three weeks later. March 2026: S&P 500 dropped 7% on the Iran war. Retail panic sold at the bottom. The index recovered to a new all-time high in 18 days. February 2026: Bitcoin dropped 15% in one day. Two days later it had bounced 11%. Same script. Every time. The crash creates the opportunity. The panic transfers the assets. The people who sell at the bottom always fund the returns of the people who don't. Here's what's actually different this time: Crypto is no longer separate from equities. Bitcoin used to be the "uncorrelated asset." That story died this week. Bitcoin moved down with the Nasdaq. Down with semis. Down with tech. Down with everything. The diversification most retail investors thought they had? They never had it. They held the same trade in three different wrappers. A Nasdaq ETF. A Bitcoin position. A pile of tech stocks. All three crashed together because all three were the same bet on cheap money. This is the moment that exposes who has a system and who was just riding a narrative. The narrative investor sees their feed full of losses and panics. They sell at the bottom. They sit in cash through the recovery. They buy back in 6 months later when the headlines feel safe again. They have already locked in the worst possible outcome. The systematic investor sees the same feed and doesn't react. Because the system already decided what to do at every price level before today happened. Surmount was built for exactly this moment. Automated, rules-based strategies that execute when the market crashes, not when your emotions do. No panic selling. No FOMO buying. Just rules. Running. While everyone else is screenshotting their losses.

BREAKING: $2.5 trillion just got wiped from global markets in 24 hours. Bitcoin crashed below $60,000. The Nasdaq dropped 4% in a single session. And every retail trader's feed is full of people posting their losses in real time... Bitcoin fell from $78,000 to $59,300 in 5 days. A 24% drop in under a week. $1.5 billion in leveraged crypto positions got liquidated in 24 hours. Long positions made up $1.28 billion of that. The fourth time in 5 days that daily liquidations crossed $1 billion. US spot Bitcoin ETFs have now posted 11 consecutive days of outflows. $3.5 billion pulled out by institutions in less than two weeks. Then the equity side cracked. Broadcom dropped 13%. Micron dropped 9.5%. SanDisk dropped 11%. Western Digital dropped 8.5%. Roughly $1 trillion in market cap erased from semiconductors alone. Strategy, the largest corporate Bitcoin holder on earth, dropped 10% after disclosing its first Bitcoin sale in years. 32 coins. $2.5 million. A rounding error against their $59 billion stack. It didn't matter. When the maximalist sells one coin, everybody else hears permission to dump everything. Now here's what actually triggered all of this. The May jobs report came in too strong. 172,000 new jobs against an 85,000 estimate. Prior months revised up by another 93,000. That single data point killed the rate cut narrative. If the economy is too strong, the Fed doesn't cut. If the Fed doesn't cut, risk assets get repriced. Crypto first. Tech stocks second. Everything else last. The entire rally of the past 12 months was built on one assumption. Rates are coming down. That assumption got vaporized in one report. And the market reacted exactly the way it always reacts: Panic. Right now Twitter is full of people posting their losses. "Down $40,000 this week." "Sold my entire crypto position." "Out of the market until things calm down." This is the exact moment retail always loses. Not because the market crashed. Because they're making decisions while watching the crash happen in real time. Every single one of these crashes follows the same pattern. October 2025: $19 billion in liquidations. The biggest wipeout in crypto history. Bitcoin made a new all-time high three weeks later. March 2026: S&P 500 dropped 7% on the Iran war. Retail panic sold at the bottom. The index recovered to a new all-time high in 18 days. February 2026: Bitcoin dropped 15% in one day. Two days later it had bounced 11%. Same script. Every time. The crash creates the opportunity. The panic transfers the assets. The people who sell at the bottom always fund the returns of the people who don't. Here's what's actually different this time: Crypto is no longer separate from equities. Bitcoin used to be the "uncorrelated asset." That story died this week. Bitcoin moved down with the Nasdaq. Down with semis. Down with tech. Down with everything. The diversification most retail investors thought they had? They never had it. They held the same trade in three different wrappers. A Nasdaq ETF. A Bitcoin position. A pile of tech stocks. All three crashed together because all three were the same bet on cheap money. This is the moment that exposes who has a system and who was just riding a narrative. The narrative investor sees their feed full of losses and panics. They sell at the bottom. They sit in cash through the recovery. They buy back in 6 months later when the headlines feel safe again. They have already locked in the worst possible outcome. The systematic investor sees the same feed and doesn't react. Because the system already decided what to do at every price level before today happened. Surmount was built for exactly this moment. Automated, rules-based strategies that execute when the market crashes, not when your emotions do. No panic selling. No FOMO buying. Just rules. Running. While everyone else is screenshotting their losses.

Surmount

48,046 views • 2 months ago

A Chinese developer created an agent system in Claude Code to sell landing pages to small businesses and, working completely solo, serves about 47 clients a month charging around $400 for each one. He built 7 agents on Claude Sonnet 4.6 capable of analyzing Google Maps in small cities, detecting businesses without websites or with totally outdated pages, and taking each opportunity all the way to a finished mockup, a promotional video, and a ready-to-send prospecting message. No assistants. No sales team. No SDRs. Just him, a MacBook, an iPhone, and an API key. While traditional agencies keep full teams to handle the same workflow, his only real costs are tokens and subscriptions to Lovable, Higgsfield, and Calendly. The 7 agents work coordinated by an orchestrator in Claude Code Router. The system consumes about 3 million tokens daily and the average API spend is just around $480 a month. They all work via MCP servers and share state using the file system, avoiding concurrency and shared memory issues. Even one of the agents lives directly on his iPhone and responds to leads while he's on the subway, in a taxi, or walking. This was the main prompt he set up: “You are the orchestrator of a solo agency that sells ready-made websites to local businesses…” The key is that the system perfectly understands what it is, what its limits are, and what goals it must achieve. It knows it has to find leads automatically. It knows it has to convert each opportunity into a landing page, a video, and a sales message without human intervention. And it knows exactly when to involve the owner. The system runs 24/7: Scout analyzes about 220 businesses daily and queues up 30 new leads. Diagnoser generates diagnostics and personalized messages for each lead. Builder creates between 3 and 5 complete landing pages for the best prospects. Filmer produces a 10-second vertical video for each proposal. Pitcher sends about 30 messages daily across 4 different channels with a response rate close to 14%. Checker automatically reviews all messages before sending them. Only when a deal exceeds $3,000 or the response rate drops below 12% does the system wake the owner. And if at that moment he's on the subway or in a taxi, the Mobile agent automatically responds to the interested lead, schedules a call in Calendly, and returns the lead to the queue. The owner just has to hit “approve” and jump into the meeting. Some real system logs: “218 businesses analyzed in Austin, Denver, and Miami. 34 without websites, 19 with 2014-era sites, and 6 with reviews requesting a redesign.” “30 messages sent. 14 responses. 5 positive. 3 Zooms scheduled.” “Landing page created for a dental clinic. Responsive. 5 sections. Video rendering.” “$3,400 agreement exceeds approved limit. Sending for manual review.” And the craziest part is that he has no dedicated servers or backend. Just a local sandbox, an MCP router, a Claude API key, and that same key connected to his iPhone. Of everything I've seen this year, it's probably the cleanest and most efficient example of a one-person automated agency: $480 a month on APIs. $18,800 in revenue. 7 prompts. A file system. And a phone in his pocket. Save this before it's too late.

A Chinese developer created an agent system in Claude Code to sell landing pages to small businesses and, working completely solo, serves about 47 clients a month charging around $400 for each one. He built 7 agents on Claude Sonnet 4.6 capable of analyzing Google Maps in small cities, detecting businesses without websites or with totally outdated pages, and taking each opportunity all the way to a finished mockup, a promotional video, and a ready-to-send prospecting message. No assistants. No sales team. No SDRs. Just him, a MacBook, an iPhone, and an API key. While traditional agencies keep full teams to handle the same workflow, his only real costs are tokens and subscriptions to Lovable, Higgsfield, and Calendly. The 7 agents work coordinated by an orchestrator in Claude Code Router. The system consumes about 3 million tokens daily and the average API spend is just around $480 a month. They all work via MCP servers and share state using the file system, avoiding concurrency and shared memory issues. Even one of the agents lives directly on his iPhone and responds to leads while he's on the subway, in a taxi, or walking. This was the main prompt he set up: “You are the orchestrator of a solo agency that sells ready-made websites to local businesses…” The key is that the system perfectly understands what it is, what its limits are, and what goals it must achieve. It knows it has to find leads automatically. It knows it has to convert each opportunity into a landing page, a video, and a sales message without human intervention. And it knows exactly when to involve the owner. The system runs 24/7: Scout analyzes about 220 businesses daily and queues up 30 new leads. Diagnoser generates diagnostics and personalized messages for each lead. Builder creates between 3 and 5 complete landing pages for the best prospects. Filmer produces a 10-second vertical video for each proposal. Pitcher sends about 30 messages daily across 4 different channels with a response rate close to 14%. Checker automatically reviews all messages before sending them. Only when a deal exceeds $3,000 or the response rate drops below 12% does the system wake the owner. And if at that moment he's on the subway or in a taxi, the Mobile agent automatically responds to the interested lead, schedules a call in Calendly, and returns the lead to the queue. The owner just has to hit “approve” and jump into the meeting. Some real system logs: “218 businesses analyzed in Austin, Denver, and Miami. 34 without websites, 19 with 2014-era sites, and 6 with reviews requesting a redesign.” “30 messages sent. 14 responses. 5 positive. 3 Zooms scheduled.” “Landing page created for a dental clinic. Responsive. 5 sections. Video rendering.” “$3,400 agreement exceeds approved limit. Sending for manual review.” And the craziest part is that he has no dedicated servers or backend. Just a local sandbox, an MCP router, a Claude API key, and that same key connected to his iPhone. Of everything I've seen this year, it's probably the cleanest and most efficient example of a one-person automated agency: $480 a month on APIs. $18,800 in revenue. 7 prompts. A file system. And a phone in his pocket. Save this before it's too late.

Marre

24,953 views • 1 month ago