Video wird geladen...

Video konnte nicht geladen werden

Beim Laden dieses Videos ist ein Problem aufgetreten. Dies könnte an einem vorübergehenden Netzwerkproblem liegen oder das Video ist möglicherweise nicht verfügbar.

"precision fix"

luqqz

1,198 subscribers

57,781 Aufrufe • vor 2 Tagen •via X (Twitter)

Wissenschaft & Technologie

Anya Rossi• Live Now

Private livecam show

0 Kommentare

Keine Kommentare verfügbar

Kommentare vom Original-Post werden hier angezeigt

Ähnliche Videos

Sheikh Ayman known for his precision in reciting and teaching, helping a young reciter fix his letter (ر) — "arra"

Sheikh Ayman known for his precision in reciting and teaching, helping a young reciter fix his letter (ر) — "arra"

deeñ

43,364 Aufrufe • vor 5 Monaten

🚀China’s slipform tech: When your irrigation channels go from hand-poured to self-leveling laser precision in one generation… 🛣️ meanwhile, the rest of the planet is still Googling ‘how to fix concrete cracks.’ 😉 📹Mechmaster

🚀China’s slipform tech: When your irrigation channels go from hand-poured to self-leveling laser precision in one generation… 🛣️ meanwhile, the rest of the planet is still Googling ‘how to fix concrete cracks.’ 😉 📹Mechmaster

Natie🇨🇳

96,909 Aufrufe • vor 1 Jahr

I hated spending minutes trying to explain which element I meant in a long prompt, so I shipped something to fix it. Inspector’s new precision UI select tool connects the element on your screen to its exact source code, saving you time on prompting.

I hated spending minutes trying to explain which element I meant in a long prompt, so I shipped something to fix it. Inspector’s new precision UI select tool connects the element on your screen to its exact source code, saving you time on prompting.

mikeBuildsMore

12,343 Aufrufe • vor 8 Monaten

The owner of this car spent thousands replacing alternator & battery in another garage but the problem was just a 50bob fix . It’s sad how small issues cost car owners because of misdiagnosis. Precision Automotive we never recommend buying new parts before we are absolutely sure of the problem

The owner of this car spent thousands replacing alternator & battery in another garage but the problem was just a 50bob fix . It’s sad how small issues cost car owners because of misdiagnosis. Precision Automotive we never recommend buying new parts before we are absolutely sure of the problem

Imran Sokwala

25,553 Aufrufe • vor 1 Jahr

BOWLS! Holy moly. I’ve just found my curling fix in the offseason. Bowls is part of the Commonwealth Games program here in Glasgow and the precision is outstanding. I nearly started sweeping the surface on some of the shots. Canada has a strong Bowls team led by Ryan Bester.

BOWLS! Holy moly. I’ve just found my curling fix in the offseason. Bowls is part of the Commonwealth Games program here in Glasgow and the precision is outstanding. I nearly started sweeping the surface on some of the shots. Canada has a strong Bowls team led by Ryan Bester.

Devin Heroux

24,910 Aufrufe • vor 10 Tagen

🚨 we’re back with the superteam india podcast! we’ve all heard of prediction markets but how do prediction markets with precision look like? on the first episode back, this episode of the superteam india podcast, Harkirat Singh sits down with Anam, blockchain engineer at Trepa — a precision prediction market built on solana (Colosseum hackathon winners by the way) where you don’t just bet yes or no, you predict exact numbers and get rewarded based on accuracy. we go deep into the technical architecture: why they store predictions in logs instead of on-chain, how rpc providers can silently truncate your program logs at 12kb, the emit cpi fix that saved them, why they rebuilt v1 from scratch over missing reserved bytes in their pdas, and the real cost challenges of sponsoring gas fees for a consumer app on solana. ⏱️ timestamps: 00:00 - cold open 00:38 - intro & guest welcome 01:08 - anam’s journey: mlh fellow → solana foundation → helius → trepa 03:04 - working at solana foundation & helius 04:20 - what are precision prediction markets? 05:03 - trepa vs polymarket: how payouts work 06:05 - the three factors: accuracy, stake & time 06:56 - what if no one else joins the pool? 08:09 - liquidity & market making challenges 09:17 - how prediction precision & steps work 10:37 - why they rebuilt v1 → v2 (reserved bytes) 12:28 - onchain storage optimization & rent reclaim 13:39 - security audit with adware labs 15:12 - the 12kb rpc log truncation discovery 15:59 - emit cpi fix: storing data as instruction data 16:33 - on-chain vs off-chain architecture deep dive 17:17 - why trepa sponsors gas (and the cost implications) 20:45 - top open source solana contracts to learn from 22:12 - what’s next: flash pools (2-minute prediction cycles) 23:38 - team size & hiring at trepa 25:50 - advice for aspiring solana smart contract devs 27:10 - outro

🚨 we’re back with the superteam india podcast! we’ve all heard of prediction markets but how do prediction markets with precision look like? on the first episode back, this episode of the superteam india podcast, Harkirat Singh sits down with Anam, blockchain engineer at Trepa — a precision prediction market built on solana (Colosseum hackathon winners by the way) where you don’t just bet yes or no, you predict exact numbers and get rewarded based on accuracy. we go deep into the technical architecture: why they store predictions in logs instead of on-chain, how rpc providers can silently truncate your program logs at 12kb, the emit cpi fix that saved them, why they rebuilt v1 from scratch over missing reserved bytes in their pdas, and the real cost challenges of sponsoring gas fees for a consumer app on solana. ⏱️ timestamps: 00:00 - cold open 00:38 - intro & guest welcome 01:08 - anam’s journey: mlh fellow → solana foundation → helius → trepa 03:04 - working at solana foundation & helius 04:20 - what are precision prediction markets? 05:03 - trepa vs polymarket: how payouts work 06:05 - the three factors: accuracy, stake & time 06:56 - what if no one else joins the pool? 08:09 - liquidity & market making challenges 09:17 - how prediction precision & steps work 10:37 - why they rebuilt v1 → v2 (reserved bytes) 12:28 - onchain storage optimization & rent reclaim 13:39 - security audit with adware labs 15:12 - the 12kb rpc log truncation discovery 15:59 - emit cpi fix: storing data as instruction data 16:33 - on-chain vs off-chain architecture deep dive 17:17 - why trepa sponsors gas (and the cost implications) 20:45 - top open source solana contracts to learn from 22:12 - what’s next: flash pools (2-minute prediction cycles) 23:38 - team size & hiring at trepa 25:50 - advice for aspiring solana smart contract devs 27:10 - outro

Superteam India

12,312 Aufrufe • vor 5 Monaten

Every tradesman or serious DIY guy who’s ever fought with a job that just won’t stay put. You’re trying to hold two pieces at a weird angle, the clamp won’t reach, the drill bit wanders, and you’re one slip away from ruining the whole run. Then you see a guy rig a ratchet strap to lock it solid, use two speed squares as a guide, or turn a damn cotton swab into a precision lube applicator. That quiet “son of a bitch… that’s brilliant” moment when a $2 fix saves you an hour and a half of cursing. I’ve been in this game long enough to know the best tools in your box are usually the ones you didn’t buy.

Every tradesman or serious DIY guy who’s ever fought with a job that just won’t stay put. You’re trying to hold two pieces at a weird angle, the clamp won’t reach, the drill bit wanders, and you’re one slip away from ruining the whole run. Then you see a guy rig a ratchet strap to lock it solid, use two speed squares as a guide, or turn a damn cotton swab into a precision lube applicator. That quiet “son of a bitch… that’s brilliant” moment when a $2 fix saves you an hour and a half of cursing. I’ve been in this game long enough to know the best tools in your box are usually the ones you didn’t buy.

FedSlayer

765,573 Aufrufe • vor 2 Monaten

hey here is the final result of octopus invaders on nvidia's flagship at full precision. nemotron super 120B on 2x H200 NVL. BF16 unquantized. 287GB of VRAM. hermes agent as the harness. 60 tok/s. first try it autonomously coded for 6 minutes straight. created 11 files. correct project structure. correct load order. started the server. i opened the browser and the result was a blank screen. i did not give up. second try i gave it a precise list of bugs and things to fix. it went back in for another 3 minutes. patched the code. served it again. still blank. so i did what any sane person would do. third try i just said the screen is blank, test it and fix it yourself. and this is where nemotron showed what it actually is. it became a debugger. you can see it in the video. realtime CSS test squares, red screen flashes, hermes agent browser tools, inspecting its own output. it built the parallax background with planets and comets. it rendered a rocket ship that tracks your mouse with fire and bullet physics. the aesthetic is real. but no enemies spawn. no collision. not playable. what surprised me is qwen 27B one shotted this exact game on a single RTX 3090 at Q4 quant. and here is nvidia's flagship at full precision on enterprise hardware needing 3 tries and still not getting there. that makes my hope high for the undisputed qwen 122B which is about to face the same test next. same hardware. same prompt and same harness. lets see if it one shots or not. full session in the video. no cuts. 5x speed.

hey here is the final result of octopus invaders on nvidia's flagship at full precision. nemotron super 120B on 2x H200 NVL. BF16 unquantized. 287GB of VRAM. hermes agent as the harness. 60 tok/s. first try it autonomously coded for 6 minutes straight. created 11 files. correct project structure. correct load order. started the server. i opened the browser and the result was a blank screen. i did not give up. second try i gave it a precise list of bugs and things to fix. it went back in for another 3 minutes. patched the code. served it again. still blank. so i did what any sane person would do. third try i just said the screen is blank, test it and fix it yourself. and this is where nemotron showed what it actually is. it became a debugger. you can see it in the video. realtime CSS test squares, red screen flashes, hermes agent browser tools, inspecting its own output. it built the parallax background with planets and comets. it rendered a rocket ship that tracks your mouse with fire and bullet physics. the aesthetic is real. but no enemies spawn. no collision. not playable. what surprised me is qwen 27B one shotted this exact game on a single RTX 3090 at Q4 quant. and here is nvidia's flagship at full precision on enterprise hardware needing 3 tries and still not getting there. that makes my hope high for the undisputed qwen 122B which is about to face the same test next. same hardware. same prompt and same harness. lets see if it one shots or not. full session in the video. no cuts. 5x speed.

Sudo su

10,994 Aufrufe • vor 4 Monaten

The world doesn't need another app that captures every "um," "like," and circular thought with 99% precision. That’s not a feature; it’s a transcript of a mess. Voice typing fast and accurately doesn't mean writing well. Standard dictation is just a high-speed way to generate garbage. If you have to spend 10 minutes editing a 2-minute voice note, the tech failed you. I got tired of my best ideas dying in my voice memos because they were too painful to polish. So we built Typeless (Typeless). It's not dictation. It's Alchemy for your thoughts. We don't just record what you said; we extract what you meant. You get the high-speed creative flow of speaking with the ruthless precision of a ghostwriter. Raw noise in, signal out. Introducing Typeless 1.0.0 for iOS & Android: - The Polish: We don't just transcribe accurately; we rewrite. Messy rants → perfect prose. - Speak to Edit: Talk back to your draft. "Make it punchier" or "Fix the hook." It actually listens. - Global Reach: Think it, speak it, ship it in any language. Private by design: - Zero data retention in the cloud - Never train on your data - On-device history storage - SOC 2, HIPAA, ISO 27001 in progress Now live on all platforms. Stop transcribing your mess. Start broadcasting your thoughts.

The world doesn't need another app that captures every "um," "like," and circular thought with 99% precision. That’s not a feature; it’s a transcript of a mess. Voice typing fast and accurately doesn't mean writing well. Standard dictation is just a high-speed way to generate garbage. If you have to spend 10 minutes editing a 2-minute voice note, the tech failed you. I got tired of my best ideas dying in my voice memos because they were too painful to polish. So we built Typeless (Typeless). It's not dictation. It's Alchemy for your thoughts. We don't just record what you said; we extract what you meant. You get the high-speed creative flow of speaking with the ruthless precision of a ghostwriter. Raw noise in, signal out. Introducing Typeless 1.0.0 for iOS & Android: - The Polish: We don't just transcribe accurately; we rewrite. Messy rants → perfect prose. - Speak to Edit: Talk back to your draft. "Make it punchier" or "Fix the hook." It actually listens. - Global Reach: Think it, speak it, ship it in any language. Private by design: - Zero data retention in the cloud - Never train on your data - On-device history storage - SOC 2, HIPAA, ISO 27001 in progress Now live on all platforms. Stop transcribing your mess. Start broadcasting your thoughts.

Huang Song

1,050,187 Aufrufe • vor 5 Monaten

A team tested Pi0, Pi0 Fast, Gr00t, and ACT on real robot arms in manufacturing tasks. (🔖 Bookmark this for later!) The task was precise: place thin rectangular frames from a messy stack into a holder. The team fine-tuned each model on 100 real trajectories and compared training time, inference speed, motion quality, and success rates. ⬇️ Here’s a breakdown of what they found Pi0 (Original) ✅ Strongest overall performance in precise pick-and-place ✅ High success rate even in edge cases ✅ Longest training time (~11 hours, ~$30 per run) ✅ Inference time of 80 ms causes short pauses between actions Despite delays, it handles complex scenarios well… solid for high-precision tasks, but slow to train. Gr00t ✅ Trains fast (~2 hours, ~$5 per run) ✅ Performs almost as well as Pi0 on large-object tasks ✅ Struggles with fine precision; random movement in some trials ✅ More training didn’t fix jitter or random offsets Best suited for tasks where exact precision isn’t critical. Not ready for manufacturing-grade accuracy without more tuning. Pi0 Fast ✅ Promised faster training, but results were underwhelming ✅ Training at 6 hours still showed low success rates ✅ Inference was slower than expected ✅ Not reliable for generalizing even slightly new tasks Currently too unstable for real-world deployment. Doesn’t live up to the “Fast” name yet. ACT (Baseline) ✅ 200MB model—lightweight, but limited ✅ Struggles with stacked objects or ambiguous scenes ✅ Success rates around 70% in best-case setups ✅ Can’t match newer models on precision or generalization Still a solid baseline, but clearly a generation behind in robustness. 🚨 Extra Notes All newer models share a common issue: •Inference takes longer than a frame (80 ms vs 33 ms), so robots “pause” between chunks. •This results in jittery movements, but not a dealbreaker unless tasks are time-sensitive. Language-conditioned tasks also fell short: after training on two labeled tasks, the model couldn’t generalize to a third unseen combination using only text prompts. ✅ The good news? These models adapt well to new robot arms with quick fine-tuning. ❌ The bad news? There’s still no plug-and-play solution for improving performance after deployment. Reinforcement learning or DAgger-style data collection during real-world operation may be the next big step, something many teams in robotics are actively working on.

A team tested Pi0, Pi0 Fast, Gr00t, and ACT on real robot arms in manufacturing tasks. (🔖 Bookmark this for later!) The task was precise: place thin rectangular frames from a messy stack into a holder. The team fine-tuned each model on 100 real trajectories and compared training time, inference speed, motion quality, and success rates. ⬇️ Here’s a breakdown of what they found Pi0 (Original) ✅ Strongest overall performance in precise pick-and-place ✅ High success rate even in edge cases ✅ Longest training time (~11 hours, ~$30 per run) ✅ Inference time of 80 ms causes short pauses between actions Despite delays, it handles complex scenarios well… solid for high-precision tasks, but slow to train. Gr00t ✅ Trains fast (~2 hours, ~$5 per run) ✅ Performs almost as well as Pi0 on large-object tasks ✅ Struggles with fine precision; random movement in some trials ✅ More training didn’t fix jitter or random offsets Best suited for tasks where exact precision isn’t critical. Not ready for manufacturing-grade accuracy without more tuning. Pi0 Fast ✅ Promised faster training, but results were underwhelming ✅ Training at 6 hours still showed low success rates ✅ Inference was slower than expected ✅ Not reliable for generalizing even slightly new tasks Currently too unstable for real-world deployment. Doesn’t live up to the “Fast” name yet. ACT (Baseline) ✅ 200MB model—lightweight, but limited ✅ Struggles with stacked objects or ambiguous scenes ✅ Success rates around 70% in best-case setups ✅ Can’t match newer models on precision or generalization Still a solid baseline, but clearly a generation behind in robustness. 🚨 Extra Notes All newer models share a common issue: •Inference takes longer than a frame (80 ms vs 33 ms), so robots “pause” between chunks. •This results in jittery movements, but not a dealbreaker unless tasks are time-sensitive. Language-conditioned tasks also fell short: after training on two labeled tasks, the model couldn’t generalize to a third unseen combination using only text prompts. ✅ The good news? These models adapt well to new robot arms with quick fine-tuning. ❌ The bad news? There’s still no plug-and-play solution for improving performance after deployment. Reinforcement learning or DAgger-style data collection during real-world operation may be the next big step, something many teams in robotics are actively working on.

Ilir Aliu

21,844 Aufrufe • vor 1 Jahr

Elon Musk says the vast majority of brain diseases are ultimately fixable with a Neuralink device "If you think of the brain like a computer, effectively like a circuit board, you can say if there were some short circuits or some circuits that should be there but aren't, we can fix those Basically, it's kind of like fixing a circuit board. The vast majority of brain issues, I think, are fixable with a Neuralink device. It’s a fine-grained way of reading and writing electrical signals in the brain at high precision If there’s an electrical storm like epilepsy, you can interrupt that storm. In the case of blindness, if somebody's lost their optic nerve or both eyes, you can still stimulate the visual cortex Anything that is a function of signals in or out, if that is the nature of the problem, it can be fixed ultimately with a Neuralink device."

Elon Musk says the vast majority of brain diseases are ultimately fixable with a Neuralink device "If you think of the brain like a computer, effectively like a circuit board, you can say if there were some short circuits or some circuits that should be there but aren't, we can fix those Basically, it's kind of like fixing a circuit board. The vast majority of brain issues, I think, are fixable with a Neuralink device. It’s a fine-grained way of reading and writing electrical signals in the brain at high precision If there’s an electrical storm like epilepsy, you can interrupt that storm. In the case of blindness, if somebody's lost their optic nerve or both eyes, you can still stimulate the visual cortex Anything that is a function of signals in or out, if that is the nature of the problem, it can be fixed ultimately with a Neuralink device."

X Freeze

17,483 Aufrufe • vor 6 Monaten

It only needs a few crazy ones to fix a continent… Let's be crazy… I got something to announce: We're launching PROTOTYPE: a new fund, fully focused on Europe. A small fund that punches way above it's size. With it we back what Europe is world-class at: robotics, automation, manufacturing, and anything that requires hard engineering. First check. First round. As early as it gets. Europe invented industry. We're the birthplace of precision manufacturing. The second largest manufacturing hub on the planet. Leaders in automation and robotics. And yet… We sell out our best tech to China. We export our best founders & most of our investment money to the US. That's insanity. What should we do instead? Build the next trillion euro companies in robotics, manufacturing, automation right here: in Europe. Showcase to young founders what is possible and change the system around them where needed. What we will do: → Publish all our fund updates and build in public: → Showcase Europe’s Most Ambitious Startups on Youtube: → Launch & support more projects like EU–INC. Enough talking about Europe. Time to build it: Startups, makerspaces, student clubs, and much more. → And most importantly, invest into the best founders in Europe. Same model as our previous funds that are in their top 1-5% cohorts worldwide. We won’t do a VC fund in the classic sense. This will be a community of hyper-ambitious people who want to actively change Europe for the better. It only needs a few crazy ones to fix a continent… Let's be crazy. Check out

It only needs a few crazy ones to fix a continent… Let's be crazy… I got something to announce: We're launching PROTOTYPE: a new fund, fully focused on Europe. A small fund that punches way above it's size. With it we back what Europe is world-class at: robotics, automation, manufacturing, and anything that requires hard engineering. First check. First round. As early as it gets. Europe invented industry. We're the birthplace of precision manufacturing. The second largest manufacturing hub on the planet. Leaders in automation and robotics. And yet… We sell out our best tech to China. We export our best founders & most of our investment money to the US. That's insanity. What should we do instead? Build the next trillion euro companies in robotics, manufacturing, automation right here: in Europe. Showcase to young founders what is possible and change the system around them where needed. What we will do: → Publish all our fund updates and build in public: → Showcase Europe’s Most Ambitious Startups on Youtube: → Launch & support more projects like EU–INC. Enough talking about Europe. Time to build it: Startups, makerspaces, student clubs, and much more. → And most importantly, invest into the best founders in Europe. Same model as our previous funds that are in their top 1-5% cohorts worldwide. We won’t do a VC fund in the classic sense. This will be a community of hyper-ambitious people who want to actively change Europe for the better. It only needs a few crazy ones to fix a continent… Let's be crazy. Check out

Andreas Klinger 🦾

401,896 Aufrufe • vor 6 Monaten

The US is pouring billions into rebuilding its semiconductor industry—but they have a problem no money can fix: they don’t have the people. 🇺🇸 By 2030, the US will face a massive shortfall of engineers, scientists, and technicians. 🇨🇳 Meanwhile, China is producing STEM grads by the millions—skilled, experienced, and dedicated. One American commented: “We searched the entire USA for Precision Engineering talent and barely filled a room. We searched in just Shanghai Province, and filled 3 football stadiums.” And a Chinese netizen wrote something that made my eyes water: “I am Chinese, with a background in science and engineering. My passion was history and literature. But my country needed tanks and cannons, so I studied physics instead. I have some sorrow—but no regrets. If my country needs physics again in my next life, I will choose it again.” This is China’s true strength. Not just the world’s most complete industrial chain, or its unmatched speed and efficiency— But a people willing to sacrifice personal dreams for the good of the nation. This is why China won’t be broken by tariffs or threats. And this is why Trump’s dream of “Made in America” is doomed to fail. Because America lacks what China has in abundance: unity, talent, and national purpose.

The US is pouring billions into rebuilding its semiconductor industry—but they have a problem no money can fix: they don’t have the people. 🇺🇸 By 2030, the US will face a massive shortfall of engineers, scientists, and technicians. 🇨🇳 Meanwhile, China is producing STEM grads by the millions—skilled, experienced, and dedicated. One American commented: “We searched the entire USA for Precision Engineering talent and barely filled a room. We searched in just Shanghai Province, and filled 3 football stadiums.” And a Chinese netizen wrote something that made my eyes water: “I am Chinese, with a background in science and engineering. My passion was history and literature. But my country needed tanks and cannons, so I studied physics instead. I have some sorrow—but no regrets. If my country needs physics again in my next life, I will choose it again.” This is China’s true strength. Not just the world’s most complete industrial chain, or its unmatched speed and efficiency— But a people willing to sacrifice personal dreams for the good of the nation. This is why China won’t be broken by tariffs or threats. And this is why Trump’s dream of “Made in America” is doomed to fail. Because America lacks what China has in abundance: unity, talent, and national purpose.

𝘊𝘰𝘳𝘳𝘪𝘯𝘦

145,157 Aufrufe • vor 1 Jahr

🇨🇳 CHINA'S MAGLEV HITS 700 KM/H IN 2 SECONDS - PLANNING 1,000 KM/H - WHILE AMERICA ARGUES ABOUT FIXING POTHOLES China just tested a maglev platform that accelerates to 700 km/h (435 mph) in 2 seconds. Target speed: 1,000 km/h (621 mph). That's faster than commercial aircraft. On the ground. The acceleration alone is borderline violent - 0 to 435 mph in two seconds is 9.8g. Fighter jet territory. Passengers would need specialized seating just to survive the launch. But let's address reality: This is a test platform. Prototype speeds don't mean operational trains. China announces ambitious projects constantly. Some materialize (their existing 430 km/h maglev in Shanghai works). Others disappear quietly. The pattern though? They're attempting scale nobody else is. High-speed rail connecting every major city. Maglev research pushed to extremes. Infrastructure spending that makes Western investment look microscopic. Meanwhile in America: Amtrak averages 105 km/h between cities. California's high-speed rail project started in 2008, burned $10+ billion, and hasn't moved a passenger. The fastest train in the U.S. hits 240 km/h for exactly one 54-mile stretch. China's going for 1,000 km/h. Even if they only achieve 800 km/h operationally, that's still triple America's maximum. Here's why this matters beyond trains: Infrastructure capacity signals industrial capability. If China can build and operate 1,000 km/h trains, they can manufacture the precision components, power systems, and control mechanisms that transfer to aerospace, military, and manufacturing. The U.S. won the 20th century partly because it built the Interstate Highway System when others couldn't. China's betting the 21st century winner will be whoever builds impossible infrastructure first. They might fail. Engineering challenges at 1,000 km/h are extreme - air resistance, track precision, emergency braking, passenger safety. But they're trying while America argues whether to fix the L train in New York. Even Chinese failure puts them ahead. You learn more from attempting the impossible than from successfully maintaining mediocrity. Source: Xinhua, CGTN

🇨🇳 CHINA'S MAGLEV HITS 700 KM/H IN 2 SECONDS - PLANNING 1,000 KM/H - WHILE AMERICA ARGUES ABOUT FIXING POTHOLES China just tested a maglev platform that accelerates to 700 km/h (435 mph) in 2 seconds. Target speed: 1,000 km/h (621 mph). That's faster than commercial aircraft. On the ground. The acceleration alone is borderline violent - 0 to 435 mph in two seconds is 9.8g. Fighter jet territory. Passengers would need specialized seating just to survive the launch. But let's address reality: This is a test platform. Prototype speeds don't mean operational trains. China announces ambitious projects constantly. Some materialize (their existing 430 km/h maglev in Shanghai works). Others disappear quietly. The pattern though? They're attempting scale nobody else is. High-speed rail connecting every major city. Maglev research pushed to extremes. Infrastructure spending that makes Western investment look microscopic. Meanwhile in America: Amtrak averages 105 km/h between cities. California's high-speed rail project started in 2008, burned $10+ billion, and hasn't moved a passenger. The fastest train in the U.S. hits 240 km/h for exactly one 54-mile stretch. China's going for 1,000 km/h. Even if they only achieve 800 km/h operationally, that's still triple America's maximum. Here's why this matters beyond trains: Infrastructure capacity signals industrial capability. If China can build and operate 1,000 km/h trains, they can manufacture the precision components, power systems, and control mechanisms that transfer to aerospace, military, and manufacturing. The U.S. won the 20th century partly because it built the Interstate Highway System when others couldn't. China's betting the 21st century winner will be whoever builds impossible infrastructure first. They might fail. Engineering challenges at 1,000 km/h are extreme - air resistance, track precision, emergency braking, passenger safety. But they're trying while America argues whether to fix the L train in New York. Even Chinese failure puts them ahead. You learn more from attempting the impossible than from successfully maintaining mediocrity. Source: Xinhua, CGTN

Mario Nawfal

818,160 Aufrufe • vor 7 Monaten

$3 Trillion is currently stuck in unpaid invoices. An average invoice takes 59 days to clear. If you're doing >$10M in revenue, getting paid in 30 days instead of 59 days will literally make you millions over 2 years. Introducing Monk. It collects your money faster and pays for itself in 30 days, guaranteed. Fast-growing companies like ElevenLabs and Profound rely on Monk. The Problem: 39% of the $3T stuck is due to two stupid reasons: 1. "Please fix this comma in the invoice and then we'll pay you'" (this happens 2-3 times per transaction) 2. "Sorry the payment reminder got buried in my inbox" (automated emails get ignored) A payment that should take 2 days, takes 10 days. Your cash on hand is embarrassingly behind "recorded revenue". To fix this, we raised $4M led by Better Tomorrow Ventures with participation from gtmfund and Anthony Danon. For our first product, we had to innovate on 3 dimensions: 1. Monk turns signed contracts into invoices with near-perfect precision - We leverage frontier models to extract key terms from your deals and turn them into invoices. 2. ⁠Monk collects payments agentically with a 24% better reply rate than automated emails: - Tuned to write emails that feel like a human request, not spam. Your invoice is competing with everything in their inbox. - We know how to find an alternate point-of-contact when someone is OOO, when to reach out, and what tonality yields a higher response rate. If you had someone on your team whose invoice emails got answered 24% more than others - that would be a big deal. You’d make them the head of revenue collection. 3. Granular visibility into your cashflow: - Understand your cash position, aging, or expansion/contraction at customer level. Book a demo at and we guarantee that Monk will pay for itself within 30 days. If you've read this far, we are giving away 40 water-tight contract templates. This would cost >$10K to any agency, startups, or freelancer, to create from scratch. Retweet this and comment 'Monk' and we'll send them to you.

$3 Trillion is currently stuck in unpaid invoices. An average invoice takes 59 days to clear. If you're doing >$10M in revenue, getting paid in 30 days instead of 59 days will literally make you millions over 2 years. Introducing Monk. It collects your money faster and pays for itself in 30 days, guaranteed. Fast-growing companies like ElevenLabs and Profound rely on Monk. The Problem: 39% of the $3T stuck is due to two stupid reasons: 1. "Please fix this comma in the invoice and then we'll pay you'" (this happens 2-3 times per transaction) 2. "Sorry the payment reminder got buried in my inbox" (automated emails get ignored) A payment that should take 2 days, takes 10 days. Your cash on hand is embarrassingly behind "recorded revenue". To fix this, we raised $4M led by Better Tomorrow Ventures with participation from gtmfund and Anthony Danon. For our first product, we had to innovate on 3 dimensions: 1. Monk turns signed contracts into invoices with near-perfect precision - We leverage frontier models to extract key terms from your deals and turn them into invoices. 2. ⁠Monk collects payments agentically with a 24% better reply rate than automated emails: - Tuned to write emails that feel like a human request, not spam. Your invoice is competing with everything in their inbox. - We know how to find an alternate point-of-contact when someone is OOO, when to reach out, and what tonality yields a higher response rate. If you had someone on your team whose invoice emails got answered 24% more than others - that would be a big deal. You’d make them the head of revenue collection. 3. Granular visibility into your cashflow: - Understand your cash position, aging, or expansion/contraction at customer level. Book a demo at and we guarantee that Monk will pay for itself within 30 days. If you've read this far, we are giving away 40 water-tight contract templates. This would cost >$10K to any agency, startups, or freelancer, to create from scratch. Retweet this and comment 'Monk' and we'll send them to you.

George Kurdin

606,168 Aufrufe • vor 9 Monaten

hey if you're thinking about running qwopus (the claude opus distilled qwen 3.5 27B) as a coding agent, this might save you a few hours. i tested both the base and the distilled version on the same hardware. single RTX 3090. same prompt. same context. same everything. the only variable was the model weights. base qwen 3.5 27B built octopus invaders in 13 minutes. 1,827 lines across 11 files. zero steering. one scope bug that took 2 lines to fix. game ran. qwopus couldn't finish the same task. enemies overlapping on screen. bullets not firing. controls worked but the game was broken. i had to steer it multiple times and it still didn't produce a playable result. both run at 35 tok/s. both use thinking mode. the distilled version actually has better jinja compatibility and doesn't stall midtask like base does on claude code. for conversation and reasoning it feels sharper. but for multifile autonomous coding where the model needs to coordinate 10+ files without losing track, base wins and it's not close. distillation compresses reasoning patterns but seems to lose precision on complex coordination. the model "thinks" well but can't hold the full picture across files the way base can. tested on opencode (base) and claude code (both). next up is hermes agent framework on base. same hardware. same prompt. comparing agents now, not just models. video below. first half is the distilled model's broken game. second half is what base built on the same 3090. judge for yourself.

hey if you're thinking about running qwopus (the claude opus distilled qwen 3.5 27B) as a coding agent, this might save you a few hours. i tested both the base and the distilled version on the same hardware. single RTX 3090. same prompt. same context. same everything. the only variable was the model weights. base qwen 3.5 27B built octopus invaders in 13 minutes. 1,827 lines across 11 files. zero steering. one scope bug that took 2 lines to fix. game ran. qwopus couldn't finish the same task. enemies overlapping on screen. bullets not firing. controls worked but the game was broken. i had to steer it multiple times and it still didn't produce a playable result. both run at 35 tok/s. both use thinking mode. the distilled version actually has better jinja compatibility and doesn't stall midtask like base does on claude code. for conversation and reasoning it feels sharper. but for multifile autonomous coding where the model needs to coordinate 10+ files without losing track, base wins and it's not close. distillation compresses reasoning patterns but seems to lose precision on complex coordination. the model "thinks" well but can't hold the full picture across files the way base can. tested on opencode (base) and claude code (both). next up is hermes agent framework on base. same hardware. same prompt. comparing agents now, not just models. video below. first half is the distilled model's broken game. second half is what base built on the same 3090. judge for yourself.

Sudo su

45,052 Aufrufe • vor 4 Monaten

Milestone! We (robotic arms for gadgets assembly) finished the first commercial order, which brought the first revenue. Here are some learnings from this: The customer was a smart toy manufacturer. The task was to add a heatsink to Raspberry Pi. We received parts from them and returned the assembled modules back. Currently, it's done by teleoperation. Later it will be done by a remote employee via the Internet. Then it will be automated action by action, reducing the operator's time on this and making the task profitable. ps. If you have an assembly task that we can do for you asynchronically - leave a comment below. Learning 1. It's possible! This task which is usually done by the human arm with 5 fingers can be done with a two-finger gripper with the addition of a couple of simple tooling. The task was not simplified. We peeled off thin films from stickers, unpacked paper boxes, moved PCB boards full of components, etc. And no unsolvable problems have been encountered yet. Challenges: 1) The paper box shifted during the opening Solved with the plastic walls that you can lean against 2) Heat pad, stuck to the gripper instead of heat sync. Can be solved by gripper with a pump, but this time solved with the patience of the operator 3) The film on the pad is very thin. Turned out that sub-millimeter arm precision is enough to peel it off with just a regular gripper. 4) The working area has not enough space. You'll only know this by doing real tasks in bulk. This could be solved by an extra pair of long arms, but in this case, solved with the patience of the operator. I think that in the end, we will have 5-10 types of universal tooling and 5-10 types of grippers to solve almost all the problems in such assembly tasks. Learning 2. It's slow. It took 5 times more time, than doing it with human hands. But the good news is there's a lot of room for improvement. We now have specific “time for task” metrics, which we will decrease with iterations. The main reasons for slowness: 1) To rotate the gripper to a steep angle you are forced to control one robot arm with two hands instead of using both arms. We can fix this by just making more room for rotations. 2) Grabbing PCB board with two arms is hard. A slight difference in rotation can break the board, and it's hard to control these angles visually. To solve this, the best way is to use force feedback so you can feel the pressure applied to the item. 3) Accuracy and steadiness is still can be improved We will try a metal version and double the motors to do this. 4) It is physically difficult for the human hands to move with such precision To solve this, we will add a pad for the hands like in surgical robots Learning 3. It's a good business model The "Factory in the cloud" is a good business model for this stage. You send us parts and we send back assembled modules. Currently, it's more convenient than sending a robot to your place, as we can iterate/fix the robot quickly and utilize it 100% of the time. When we polish the set-up over time - we can send robots to your place. So if we can assemble something for you in the USA with Chinese prices by using modern automation - leave a comment below.

Milestone! We (robotic arms for gadgets assembly) finished the first commercial order, which brought the first revenue. Here are some learnings from this: The customer was a smart toy manufacturer. The task was to add a heatsink to Raspberry Pi. We received parts from them and returned the assembled modules back. Currently, it's done by teleoperation. Later it will be done by a remote employee via the Internet. Then it will be automated action by action, reducing the operator's time on this and making the task profitable. ps. If you have an assembly task that we can do for you asynchronically - leave a comment below. Learning 1. It's possible! This task which is usually done by the human arm with 5 fingers can be done with a two-finger gripper with the addition of a couple of simple tooling. The task was not simplified. We peeled off thin films from stickers, unpacked paper boxes, moved PCB boards full of components, etc. And no unsolvable problems have been encountered yet. Challenges: 1) The paper box shifted during the opening Solved with the plastic walls that you can lean against 2) Heat pad, stuck to the gripper instead of heat sync. Can be solved by gripper with a pump, but this time solved with the patience of the operator 3) The film on the pad is very thin. Turned out that sub-millimeter arm precision is enough to peel it off with just a regular gripper. 4) The working area has not enough space. You'll only know this by doing real tasks in bulk. This could be solved by an extra pair of long arms, but in this case, solved with the patience of the operator. I think that in the end, we will have 5-10 types of universal tooling and 5-10 types of grippers to solve almost all the problems in such assembly tasks. Learning 2. It's slow. It took 5 times more time, than doing it with human hands. But the good news is there's a lot of room for improvement. We now have specific “time for task” metrics, which we will decrease with iterations. The main reasons for slowness: 1) To rotate the gripper to a steep angle you are forced to control one robot arm with two hands instead of using both arms. We can fix this by just making more room for rotations. 2) Grabbing PCB board with two arms is hard. A slight difference in rotation can break the board, and it's hard to control these angles visually. To solve this, the best way is to use force feedback so you can feel the pressure applied to the item. 3) Accuracy and steadiness is still can be improved We will try a metal version and double the motors to do this. 4) It is physically difficult for the human hands to move with such precision To solve this, we will add a pad for the hands like in surgical robots Learning 3. It's a good business model The "Factory in the cloud" is a good business model for this stage. You send us parts and we send back assembled modules. Currently, it's more convenient than sending a robot to your place, as we can iterate/fix the robot quickly and utilize it 100% of the time. When we polish the set-up over time - we can send robots to your place. So if we can assemble something for you in the USA with Chinese prices by using modern automation - leave a comment below.

Igor Kulakov

37,266 Aufrufe • vor 1 Jahr

Elon Musk is building the first system in history that makes the human body optional. He said it out loud. Nobody heard what he was actually saying. Neuralink reads the brain. Optimus builds the body. Two companies. One architecture. The world covered them as separate stories. Musk: “Combine that with a Neuralink.” One system decodes neural signals. The other provides hardware that executes them. Thought becomes superhuman motion. No nerve delay. No muscle ceiling. No biological decay. Musk: “The motor commands from your brain… now go to your robot arms or robot legs.” Your brain fires. Robotic hardware answers with force and precision flesh never had. Musk: “You’d have basically Cybernetics Superpowers.” Not metaphor. Engineering specification. Limbs that never fatigue. Never weaken. React faster than reflexes allow. Lift what biology cannot. Amputees get access first. Medical necessity opens the door. But once enhanced humans measurably outperform biological ones, the entire meaning of disability inverts. Disability stops being what you fix. It becomes what you already have. You’re running the body you were born into. Not because it’s optimal. Because it was the only hardware available. The moment an upgrade integrates directly with your brain, keeping the original isn’t preservation. It’s self-imposed limitation. Your body isn’t you. It’s the first hardware your consciousness was ever loaded onto. Musk is building the second. Nobody chooses inferior equipment when superior equipment obeys their thoughts. That’s not speculation. That’s human nature operating exactly as it always has. Musk didn’t stumble into this. He engineered it from the start. Neuralink was never a side project. Optimus was never a distraction. Two halves of one system designed to make biology a choice, not a constraint. Everyone covered two companies. Two missions. Two timelines. Musk was building one product the entire time. He told you. You just weren’t listening.

Elon Musk is building the first system in history that makes the human body optional. He said it out loud. Nobody heard what he was actually saying. Neuralink reads the brain. Optimus builds the body. Two companies. One architecture. The world covered them as separate stories. Musk: “Combine that with a Neuralink.” One system decodes neural signals. The other provides hardware that executes them. Thought becomes superhuman motion. No nerve delay. No muscle ceiling. No biological decay. Musk: “The motor commands from your brain… now go to your robot arms or robot legs.” Your brain fires. Robotic hardware answers with force and precision flesh never had. Musk: “You’d have basically Cybernetics Superpowers.” Not metaphor. Engineering specification. Limbs that never fatigue. Never weaken. React faster than reflexes allow. Lift what biology cannot. Amputees get access first. Medical necessity opens the door. But once enhanced humans measurably outperform biological ones, the entire meaning of disability inverts. Disability stops being what you fix. It becomes what you already have. You’re running the body you were born into. Not because it’s optimal. Because it was the only hardware available. The moment an upgrade integrates directly with your brain, keeping the original isn’t preservation. It’s self-imposed limitation. Your body isn’t you. It’s the first hardware your consciousness was ever loaded onto. Musk is building the second. Nobody chooses inferior equipment when superior equipment obeys their thoughts. That’s not speculation. That’s human nature operating exactly as it always has. Musk didn’t stumble into this. He engineered it from the start. Neuralink was never a side project. Optimus was never a distraction. Two halves of one system designed to make biology a choice, not a constraint. Everyone covered two companies. Two missions. Two timelines. Musk was building one product the entire time. He told you. You just weren’t listening.

Dustin

17,271 Aufrufe • vor 20 Tagen

$Tired of staring at the "backstage" of your house? There’s a classic homeowner dilemma: you need the utilities to live comfortably, but you don't necessarily want to see the electric meter and AC unit every time you step into the backyard. One man decided to tackle this "hideous eyesore" head-on by building a high-end privacy screen that proves utility can actually be beautiful. This isn't just your average DIY fence. From digging deep post holes (and dodging irrigation pipes) to mixing "cake batter from hell," every step was a lesson in precision. After securing the 4 \times 4 posts in 100 lbs of concrete, the real magic happened with a sleek color palette. The posts were primed and painted in Sherwin Williams’ "Tricorn Black"—the blackest black available—to create a modern, industrial foundation. To keep things natural yet refined, he opted for cedar slats. But because the boards were a bit "furry," they got the VIP treatment through a planer for a smooth-as-glass finish. The final touches? Cedar Stain: To bring out those warm, organic tones. Custom Corner Trim: Painted black to match the posts, elevating the project from a "weekend fix" to a professional-grade architectural feature. Solar Post Caps: Adding a subtle glow when the sun goes down. The result is a stunning contrast of black and cedar that perfectly masks the home's "vital organs." And for the critics of the color combo? The master plan involves a white lime-wash for the red brick, ensuring this screen will truly pop. Sometimes, the best way to handle an eyesore is to give it a stylish place to hide. Who else is ready to start their next backyard renovation?$

Tired of staring at the "backstage" of your house? There’s a classic homeowner dilemma: you need the utilities to live comfortably, but you don't necessarily want to see the electric meter and AC unit every time you step into the backyard. One man decided to tackle this "hideous eyesore" head-on by building a high-end privacy screen that proves utility can actually be beautiful. This isn't just your average DIY fence. From digging deep post holes (and dodging irrigation pipes) to mixing "cake batter from hell," every step was a lesson in precision. After securing the 4 \times 4 posts in 100 lbs of concrete, the real magic happened with a sleek color palette. The posts were primed and painted in Sherwin Williams’ "Tricorn Black"—the blackest black available—to create a modern, industrial foundation. To keep things natural yet refined, he opted for cedar slats. But because the boards were a bit "furry," they got the VIP treatment through a planer for a smooth-as-glass finish. The final touches? Cedar Stain: To bring out those warm, organic tones. Custom Corner Trim: Painted black to match the posts, elevating the project from a "weekend fix" to a professional-grade architectural feature. Solar Post Caps: Adding a subtle glow when the sun goes down. The result is a stunning contrast of black and cedar that perfectly masks the home's "vital organs." And for the critics of the color combo? The master plan involves a white lime-wash for the red brick, ensuring this screen will truly pop. Sometimes, the best way to handle an eyesore is to give it a stylish place to hide. Who else is ready to start their next backyard renovation?

PeachProof

27,728 Aufrufe • vor 2 Monaten

$Elon Musk just said the blind will see better than anyone with working eyes. That’s not optimism. That’s a spec sheet. Musk: “And then over time, I think you get to higher resolution than human eyes.” Not restoring vision. Making it obsolete. Human eyes peaked millions of years ago. Digital input has no peak. Musk: “So like Geordi LaForge from Star Trek.” The VISOR didn’t fix eyesight. It accessed the entire electromagnetic spectrum. That’s the blueprint. Musk: “Do you want to see in radar? No problem.” Vision stops being what your retina allows. Becomes what you select. Musk: “You can see ultraviolet, infrared, eagle vision, whatever you want. You could also see in different wavelengths.” Heat signatures. Microscopic precision. Wavelengths your eyes were never built to detect because survival never required them. All toggleable. Resolution with no biological ceiling. Blindness stops being a deficit. Becomes early access. First Neuralink patients won’t be recovering what they lost. They’ll be accessing what no one has ever had. Intact biology becomes the limitation. Every human who has ever lived has seen the same narrow fraction of what actually exists. Same wavelengths. Same resolution. Same biological ceiling for 300,000 years. Everything humanity has ever built, discovered, or understood, we did while perceiving less than one percent of what’s actually there. We called that vision. It was just enough not to die. You’ve never seen the world. You’ve seen your body’s opinion of it. Neuralink doesn’t improve that system. It replaces it. Two people will stand in the same room. One with biological eyes. One with Neuralink. They won’t be looking at the same reality. The blind won’t just see again. They’ll see more than anyone reading this. And for the first time in 300,000 years, the rest of us will know exactly how little we’ve been seeing.$

Elon Musk just said the blind will see better than anyone with working eyes. That’s not optimism. That’s a spec sheet. Musk: “And then over time, I think you get to higher resolution than human eyes.” Not restoring vision. Making it obsolete. Human eyes peaked millions of years ago. Digital input has no peak. Musk: “So like Geordi LaForge from Star Trek.” The VISOR didn’t fix eyesight. It accessed the entire electromagnetic spectrum. That’s the blueprint. Musk: “Do you want to see in radar? No problem.” Vision stops being what your retina allows. Becomes what you select. Musk: “You can see ultraviolet, infrared, eagle vision, whatever you want. You could also see in different wavelengths.” Heat signatures. Microscopic precision. Wavelengths your eyes were never built to detect because survival never required them. All toggleable. Resolution with no biological ceiling. Blindness stops being a deficit. Becomes early access. First Neuralink patients won’t be recovering what they lost. They’ll be accessing what no one has ever had. Intact biology becomes the limitation. Every human who has ever lived has seen the same narrow fraction of what actually exists. Same wavelengths. Same resolution. Same biological ceiling for 300,000 years. Everything humanity has ever built, discovered, or understood, we did while perceiving less than one percent of what’s actually there. We called that vision. It was just enough not to die. You’ve never seen the world. You’ve seen your body’s opinion of it. Neuralink doesn’t improve that system. It replaces it. Two people will stand in the same room. One with biological eyes. One with Neuralink. They won’t be looking at the same reality. The blind won’t just see again. They’ll see more than anyone reading this. And for the first time in 300,000 years, the rest of us will know exactly how little we’ve been seeing.

Dustin

32,562 Aufrufe • vor 12 Tagen