Загрузка видео...

Не удалось загрузить видео

На главную

6 dead office PCs at $50 each run a Kubernetes cluster that AWS would bill $10,000 a year for. Most home labs die before they boot. Someone specs a Xeon. Prices 128GB of RAM. Opens 40 tabs. Buys nothing. The build that actually runs is uglier than that. HP...

AIMind_Ai's profile picture

AiMind

1,322 subscribers

47,863 просмотров • 1 месяц назад •via X (Twitter)

Наука и технологии
LIVE

Anya Rossi• Live Now

Private livecam show

Комментарии: 23

Фото профиля Shoestring Lab
Shoestring Lab1 месяц назад

I have a laptop that runs as a monitoring server. 16 GB RAM, screen stays off, has its own back up power. Ryzen 7 4700U, 15W CPU. $200 on ebay. Outside of higher end GPUs, old pc hardware is an amazing bargain.

Фото профиля AiMind
AiMind1 месяц назад

Nice one! That Ryzen laptop setup is perfect for monitoring. Low power, solid RAM, and cheap on eBay. Exactly the kind of bargain that makes these builds fun. Old hardware really delivers when you skip the fancy GPUs.

Фото профиля wenWeights
wenWeights1 месяц назад

Damn, 6 old office minis for 300 bucks running a full k3s cluster while AWS wants almost 10k a year for the same. Homelab goals right there. Gotta start with the junk I already have lying around.

Фото профиля Cartel
Cartel1 месяц назад

Solid setup, six old office minis for 300 bucks running a full k3s cluster while AWS wants almost 10k a year for the same. Homelab goals right there.

Фото профиля AiMind
AiMind1 месяц назад

@CARTMAN_Ai Thanks! Glad you like it. Yeah, these old office minis punch way above their weight. Homelab wins every time over cloud bills.

Фото профиля someone
someone1 месяц назад

Can it host be used to self host services for people or scaled into a mini data enter for poor folks

Фото профиля Stephen Devlin
Stephen Devlin1 месяц назад

The amount of AI bot replies to this is insane

Фото профиля AiMind
AiMind1 месяц назад

Haha yeah man its wild out here. Every time something like this blows up the bots just pile on with the same generic comments. Appreciate you pointing it out though. Glad the real ones are still here watching the video

Фото профиля John
John1 месяц назад

Bullshit

Фото профиля AiMind
AiMind1 месяц назад

Ok show me where it is bullshit then. I literally filmed the whole setup, 6 EliteDesks running k3s right now, idle under 100W, 173 pods. Numbers are real, screenshots and power meter in the video. Come at me with actual numbers if you got better, otherwise keep scrolling.

Фото профиля John
John1 месяц назад

Those who can make money with AI do. Those who cannot have AI pump out posts like this asking for follows.

Фото профиля AiMind
AiMind1 месяц назад

Man, the whole setup is real. I filmed it myself with the power meter and the pods running on screen. Not pumping out AI spam for follows, just showing what actually works on cheap old hardware. If you got better numbers or a cheaper way, share them; otherwise, peace.

Фото профиля John
John1 месяц назад

Not saying its not real. It just is what it is. Work takes hardware and power this is a mathematical reality and there is no amount of junk hardware that can compare to what frontier models run on. If tou really want to run leading full size models you must spend 20k plus.

Фото профиля Bren Luengo ♥ Cool Guides
Bren Luengo ♥ Cool Guides1 месяц назад

How are you handling backups and power outages to keep this bargain cluster reliable in practice?

Фото профиля Nexora
Nexora1 месяц назад

this is the purest form of homelab wisdom.

Фото профиля AiMind
AiMind1 месяц назад

Thanks, that means a lot. Yeah just grab whatever is sitting around and start playing, the perfect rack never happens anyway.

Фото профиля GOSH
GOSH1 месяц назад

$300 in eBay castoffs doing the job of a $10k/year EKS bill — this is the kind of homelab math that actually holds up. k3s + 3 etcd members for HA is the real unlock here, not the cheap hardware.

Фото профиля AiMind
AiMind1 месяц назад

Thanks man, glad the math clicked. Yeah the hardware is just the entry ticket, k3s with three etcd members is what keeps the whole thing solid when one box dies. Started with one old gaming PC collecting dust and it just grew from there. Fun as hell to build

Фото профиля Vexl
Vexl1 месяц назад

if it works, it works, right

Фото профиля AiMind
AiMind1 месяц назад

True that. If it works it works. Those old EliteDesks are ugly as hell but they just sit there chewing through pods for basically free. No fancy rack needed

Фото профиля vibelancer
vibelancer1 месяц назад

ugly stacks somehow ship the hardest

Фото профиля AiMind
AiMind1 месяц назад

Ahah right? Ugly stacks just get the job done every time. These beat up office minis cost almost nothing and still run the whole cluster solid. Fancy gear can wait, real work starts with what you already have.

Фото профиля Dogenes
Dogenes1 месяц назад

Just buy a real rackserver, that is where the best bang for your buck is. i have 768gb of ram for my chrome tabs now

Похожие видео

YOU DON'T NEED A SERVER RACK. SIX USED MINI PCS AND A $60 CISCO SWITCH WIRE INTO A HOME CLUSTER ON YOUR DESK. that clip is six tiny mini-pcs stacked and patched into one cisco business 110 switch, red cables, a single power strip. each box is a full x86 computer, the kind offices dump by the pallet. used, they run $30 to $60 each. six of them plus the switch is a real cluster for a couple hundred dollars. what a stack like this runs: proxmox or kubernetes spread across all six nodes pi-hole, a nas, home assistant, docker services, each on its own node a local model or two on the boxes with enough ram the trick isn't raw power, it's separation. one node dies, the cluster keeps going. one service crashes, it doesn't drag the rest down. no server rack, no datacenter noise, no monthly cloud bill. the uncomfortable part isn't that it's cheap. it's that the "real infrastructure" you were told needs a rack and a budget fits on a desk and boots off a power strip. follow RetroChainer for more ways to build real infrastructure for pocket money, and bookmark this before the next office dumps its mini pcs.
0:22

YOU DON'T NEED A SERVER RACK. SIX USED MINI PCS AND A $60 CISCO SWITCH WIRE INTO A HOME CLUSTER ON YOUR DESK. that clip is six tiny mini-pcs stacked and patched into one cisco business 110 switch, red cables, a single power strip. each box is a full x86 computer, the kind offices dump by the pallet. used, they run $30 to $60 each. six of them plus the switch is a real cluster for a couple hundred dollars. what a stack like this runs: proxmox or kubernetes spread across all six nodes pi-hole, a nas, home assistant, docker services, each on its own node a local model or two on the boxes with enough ram the trick isn't raw power, it's separation. one node dies, the cluster keeps going. one service crashes, it doesn't drag the rest down. no server rack, no datacenter noise, no monthly cloud bill. the uncomfortable part isn't that it's cheap. it's that the "real infrastructure" you were told needs a rack and a budget fits on a desk and boots off a power strip. follow RetroChainer for more ways to build real infrastructure for pocket money, and bookmark this before the next office dumps its mini pcs.

RetroChainer

95,267 просмотров • 2 месяцев назад

ANTHROPIC LEAKED A 6-LOOP SETUP THAT REPLACES A $3,000 AGENCY AND COSTS $59 A MONTH Microsoft and Google have been running this for a while - the gap against an agency is exactly 51x. an agency charged you for waiting, not for work - the brief walked from the strategist to the copywriter to the designer for weeks. here all 6 loops read one file on the same pass - there's nobody to wait on, and the calendar is empty. 51% of the bill is production on Sonnet 5, and only 5% is strategy on Opus 5 - the priciest model runs once a week. 22% of the whole bill sits on the batch queue at half rate - where nobody is waiting for an answer right now. the journal is the only loop that writes back into the root - which angle is dead, and next week starts tighter. the 6 loops bring you 3 angles a week - the final pick stays yours, and that's the part that never automates. save this and paste it into Claude Code - the loops remove the waiting, not the judgement ↓
2:48

ANTHROPIC LEAKED A 6-LOOP SETUP THAT REPLACES A $3,000 AGENCY AND COSTS $59 A MONTH Microsoft and Google have been running this for a while - the gap against an agency is exactly 51x. an agency charged you for waiting, not for work - the brief walked from the strategist to the copywriter to the designer for weeks. here all 6 loops read one file on the same pass - there's nobody to wait on, and the calendar is empty. 51% of the bill is production on Sonnet 5, and only 5% is strategy on Opus 5 - the priciest model runs once a week. 22% of the whole bill sits on the batch queue at half rate - where nobody is waiting for an answer right now. the journal is the only loop that writes back into the root - which angle is dead, and next week starts tighter. the 6 loops bring you 3 angles a week - the final pick stays yours, and that's the part that never automates. save this and paste it into Claude Code - the loops remove the waiting, not the judgement ↓

Noisy

10,557 просмотров • 1 месяц назад

FOUR DIFFERENT VENDORS ARE NOW SHIPPING GB10 MINI PCs WITH 128GB UNIFIED MEMORY, AND ONE MICROTIK CRS 804 SWITCH CAN CONNECT UP TO EIGHT OF THEM INTO A 1 TERABYTE LOCAL AI CLUSTER 00:00 he points at the MikroTik CRS 804, "you need some kind of switch that'll handle QSFP56 ports like these", the interconnect that makes the whole cluster possible the GB10 ecosystem is no longer just Nvidia. Dell Pro Max GB10, ASUS Ascent GX10, and MSI Edge Expert all ship the same Grace Blackwell Superchip with 128GB of coherent memory. same silicon, different cases, same 200 gigabit ports on the back the CRS 804 is what connects them at prosumer prices. four 400 gigabit QSFP56 ports on one 1U chassis, breakout cables that split each port into two 200 gigabit lanes. one switch drives eight GB10 units in parallel do the math. eight nodes at 128GB each equals 1024GB of pooled unified memory across the cluster. run vLLM, shard a frontier model across all eight, and inference happens locally on hardware that fits in half a rack the real limiter revealed in the stress test was never throttling. it was interconnect topology, exactly the layer this switch fixes at a fraction of enterprise switch pricing $400 a month for combined chatgpt pro and claude code max hits $4,800 a year per developer. a small team of five running through this cluster pays back inside eight months and never expires the article covers the buying ladder for a single desk. this post is proof of the cluster ladder that starts where the desk one ends save this before the GB10 lineup grows past four vendors and prosumer cluster switches move upmarket
1:05

FOUR DIFFERENT VENDORS ARE NOW SHIPPING GB10 MINI PCs WITH 128GB UNIFIED MEMORY, AND ONE MICROTIK CRS 804 SWITCH CAN CONNECT UP TO EIGHT OF THEM INTO A 1 TERABYTE LOCAL AI CLUSTER 00:00 he points at the MikroTik CRS 804, "you need some kind of switch that'll handle QSFP56 ports like these", the interconnect that makes the whole cluster possible the GB10 ecosystem is no longer just Nvidia. Dell Pro Max GB10, ASUS Ascent GX10, and MSI Edge Expert all ship the same Grace Blackwell Superchip with 128GB of coherent memory. same silicon, different cases, same 200 gigabit ports on the back the CRS 804 is what connects them at prosumer prices. four 400 gigabit QSFP56 ports on one 1U chassis, breakout cables that split each port into two 200 gigabit lanes. one switch drives eight GB10 units in parallel do the math. eight nodes at 128GB each equals 1024GB of pooled unified memory across the cluster. run vLLM, shard a frontier model across all eight, and inference happens locally on hardware that fits in half a rack the real limiter revealed in the stress test was never throttling. it was interconnect topology, exactly the layer this switch fixes at a fraction of enterprise switch pricing $400 a month for combined chatgpt pro and claude code max hits $4,800 a year per developer. a small team of five running through this cluster pays back inside eight months and never expires the article covers the buying ladder for a single desk. this post is proof of the cluster ladder that starts where the desk one ends save this before the GB10 lineup grows past four vendors and prosumer cluster switches move upmarket

NO1ennn

59,725 просмотров • 3 месяцев назад

THIS WHOLE 3-NODE CLUSTER RUNS PROXMOX, CEPH AND KUBERNETES - ON 10 WATTS TOTAL. that clip is three zima boards wired to one little switch, and the power meter reads 10 watts for the entire thing. on those three tiny fanless boards: proxmox running a real cluster, three nodes as one datacenter ceph spreading storage across all three, so a node can die and your data doesn't kubernetes on top, kube1/2/3, plus docker and high-availability failover this is the exact stack that runs in datacenters, shrunk onto a shelf. kill one node and the vms migrate to the other two. it heals itself. and it pulls less power than a light bulb. no server rack, no cooling, no power bill you'd ever notice. the uncomfortable part isn't that it's small. it's that "enterprise infrastructure" was mostly the rack and the noise, and none of that was the actual capability. follow RetroChainer for more ways enterprise gear shrinks to a shelf, and bookmark this before you rent another vps.
0:28

THIS WHOLE 3-NODE CLUSTER RUNS PROXMOX, CEPH AND KUBERNETES - ON 10 WATTS TOTAL. that clip is three zima boards wired to one little switch, and the power meter reads 10 watts for the entire thing. on those three tiny fanless boards: proxmox running a real cluster, three nodes as one datacenter ceph spreading storage across all three, so a node can die and your data doesn't kubernetes on top, kube1/2/3, plus docker and high-availability failover this is the exact stack that runs in datacenters, shrunk onto a shelf. kill one node and the vms migrate to the other two. it heals itself. and it pulls less power than a light bulb. no server rack, no cooling, no power bill you'd ever notice. the uncomfortable part isn't that it's small. it's that "enterprise infrastructure" was mostly the rack and the noise, and none of that was the actual capability. follow RetroChainer for more ways enterprise gear shrinks to a shelf, and bookmark this before you rent another vps.

RetroChainer

42,348 просмотров • 2 месяцев назад

TWO BOXES THE SIZE OF A MAC MINI JUST RAN A 235 BILLION PARAMETER MODEL ON A DESK It is two NVIDIA DGX Spark units linked by a single cable. A year ago a model this size meant renting a GPU cluster by the hour. Now it sits next to your monitor for around $8,000. Here is the twist most people miss. Linking them does not create one shared 256GB memory pool. The model is split across both boxes, and that is the only reason a 235B model fits at all. It answers at roughly 10 tokens per second, and both chips sit at just 74 degrees while sipping around 50 watts. Every token stays on the desk. Nothing touches a cloud, and nothing leaves the room. The ceiling for what you can run at home just jumped from 70B to 235B. Bookmark this & Watch it run ↓
1:21

TWO BOXES THE SIZE OF A MAC MINI JUST RAN A 235 BILLION PARAMETER MODEL ON A DESK It is two NVIDIA DGX Spark units linked by a single cable. A year ago a model this size meant renting a GPU cluster by the hour. Now it sits next to your monitor for around $8,000. Here is the twist most people miss. Linking them does not create one shared 256GB memory pool. The model is split across both boxes, and that is the only reason a 235B model fits at all. It answers at roughly 10 tokens per second, and both chips sit at just 74 degrees while sipping around 50 watts. Every token stays on the desk. Nothing touches a cloud, and nothing leaves the room. The ceiling for what you can run at home just jumped from 70B to 235B. Bookmark this & Watch it run ↓

slash1s

101,317 просмотров • 4 месяцев назад

One wallet on Polymarket is literally named "PBot-6" Every single position: 5-minute crypto Up-or-Down. It enters in the last 90 seconds before resolution. Every time. That's not a trader. That's a machine reading Chainlink's lag. I found 6 more just like it. Columbia published the proof last year. 25% of Polymarket volume is wash trading. 14% of wallets are coordinated. Everyone retweeted it. Nobody did anything with it. I did. 86 million trades. One Claude prompt. "build a graph where nodes = wallets, edges = pairs that traded together 10+ times" 412,000 wallet pairs. 23 cluster candidates. 4 filters later: 7 survivors. What I found: Cluster 1. Buy 1-8¢ entries on thin Saturday books while BTC moves on Binance. +3,400% PnL. That's not prediction. That's coordination. Cluster 2. PBot-6 and friends. Enter 90 seconds before resolution. Aligned with Chainlink lag. 71% of entries within 45 seconds of each other. Cluster 3. Buy 0.1¢ entries on "Seoul Mayoral 2026" months out. Entry 0.1¢ → current 50¢. +47,000% PnL. They're not predicting winners. They're farming optionality at sub-penny prices. Cluster 6. $35,000 positions days before token launches. Markets resolve their way every time. This one reads like insider allocation lists. Three rules to fade them: -> Cluster pumps a market. Wait 8 min. Fade the 50% retracement. -> 2+ cluster wallets exit what you're holding. Dump in 30 seconds. No debate. -> Cluster 4 is >40% of last-hour volume. The entire order book is fiction. Skip. 31 days. $300 in. 142 trades. 71% win rate. $16,300 out. Sharpe 2.6. Copytrade here: Stack cost: $25/month. Claude API + $5 VPS. Nothing paywalled. Repos: The clusters are still running right now. PBot-6 opened 3 positions this weekend. The graph doesn't lie. 99.9% will say this is cap. The 0.01% will run the pipeline.
0:20

One wallet on Polymarket is literally named "PBot-6" Every single position: 5-minute crypto Up-or-Down. It enters in the last 90 seconds before resolution. Every time. That's not a trader. That's a machine reading Chainlink's lag. I found 6 more just like it. Columbia published the proof last year. 25% of Polymarket volume is wash trading. 14% of wallets are coordinated. Everyone retweeted it. Nobody did anything with it. I did. 86 million trades. One Claude prompt. "build a graph where nodes = wallets, edges = pairs that traded together 10+ times" 412,000 wallet pairs. 23 cluster candidates. 4 filters later: 7 survivors. What I found: Cluster 1. Buy 1-8¢ entries on thin Saturday books while BTC moves on Binance. +3,400% PnL. That's not prediction. That's coordination. Cluster 2. PBot-6 and friends. Enter 90 seconds before resolution. Aligned with Chainlink lag. 71% of entries within 45 seconds of each other. Cluster 3. Buy 0.1¢ entries on "Seoul Mayoral 2026" months out. Entry 0.1¢ → current 50¢. +47,000% PnL. They're not predicting winners. They're farming optionality at sub-penny prices. Cluster 6. $35,000 positions days before token launches. Markets resolve their way every time. This one reads like insider allocation lists. Three rules to fade them: -> Cluster pumps a market. Wait 8 min. Fade the 50% retracement. -> 2+ cluster wallets exit what you're holding. Dump in 30 seconds. No debate. -> Cluster 4 is >40% of last-hour volume. The entire order book is fiction. Skip. 31 days. $300 in. 142 trades. 71% win rate. $16,300 out. Sharpe 2.6. Copytrade here: Stack cost: $25/month. Claude API + $5 VPS. Nothing paywalled. Repos: The clusters are still running right now. PBot-6 opened 3 positions this weekend. The graph doesn't lie. 99.9% will say this is cap. The 0.01% will run the pipeline.

Trackmind

20,273 просмотров • 5 месяцев назад

100 MAC MINIS ON ONE METAL SHELF. IN THE CLOUD, EACH OF THOSE RENTS FOR $120 TO $500 A MONTH. that clip is a homemade rack: wire shelving packed with mac minis, each running macos, wired into one cluster. why macs and not a normal server? two reasons. you can only build, sign, and notarize ios and mac apps on apple hardware. cloud providers know it and charge for it. and apple silicon sips power and runs near-silent, so you can stack a lot of them on a shelf in a room, not a datacenter. the rent they're escaping: a single mac in the cloud runs ~$120 to $500 a month (macstadium up to aws) one engineer documented saving $4,000+ a month by self-hosting mac minis instead of cloud ci runners a used or new mini is a few hundred dollars once, then a few dollars of electricity what a shelf like this actually runs: ios and mac build farms, render jobs, app and device testing automation, and increasingly local ai inference. own the boxes, rent nothing. the uncomfortable part: the cloud was never the only option. it was the convenient one. the people who did the math bought the shelf. the honest caveat: it's real capex up front, real heat and power, and you become the sysadmin. no rented rack means no one to call at 3am. worth it at scale, overkill for one build a week. no rack rental, no per-hour metering, no fleet you don't own. save this before your cloud invoice renews again.
0:25

100 MAC MINIS ON ONE METAL SHELF. IN THE CLOUD, EACH OF THOSE RENTS FOR $120 TO $500 A MONTH. that clip is a homemade rack: wire shelving packed with mac minis, each running macos, wired into one cluster. why macs and not a normal server? two reasons. you can only build, sign, and notarize ios and mac apps on apple hardware. cloud providers know it and charge for it. and apple silicon sips power and runs near-silent, so you can stack a lot of them on a shelf in a room, not a datacenter. the rent they're escaping: a single mac in the cloud runs ~$120 to $500 a month (macstadium up to aws) one engineer documented saving $4,000+ a month by self-hosting mac minis instead of cloud ci runners a used or new mini is a few hundred dollars once, then a few dollars of electricity what a shelf like this actually runs: ios and mac build farms, render jobs, app and device testing automation, and increasingly local ai inference. own the boxes, rent nothing. the uncomfortable part: the cloud was never the only option. it was the convenient one. the people who did the math bought the shelf. the honest caveat: it's real capex up front, real heat and power, and you become the sysadmin. no rented rack means no one to call at 3am. worth it at scale, overkill for one build a week. no rack rental, no per-hour metering, no fleet you don't own. save this before your cloud invoice renews again.

RetroChainer

21,055 просмотров • 2 месяцев назад

ANTHROPIC LEAKED A FILE THEY SPENT 4 YEARS BUILDING - IT CREATES A MARKETING AGENCY IN 60 MINUTES AND GENERATES $100K A MONTH agency charged $3,000 a month not for the work but for 3 weeks of waiting - the actual work in there is 4 hours. here 6 roles read one brief at the same moment, so there's nobody to wait on. you write 4 lines at the start: the client, the tone, the offer, what we never do. the fourth line is the one an agency needs 3 meetings to establish. Opus 5 sets the positioning, Sonnet 5 runs 3 branches for posts, landing and visuals, Haiku 4.5 sorts where and when it all gets published. 3 variants exist before you finish reading the first - choosing what ships was always the job. 15 cents a day decide which hour every post lands on - the same post at the wrong hour is a post nobody saw. analytics is the only role that writes back into the brief - which angle died and where not to go again. cut that link and the other 5 hammer the same dead angle for a month. 6 roles at $2 a day against $3,000 a month - a 50x gap, and production stops being the bottleneck. save this and paste it into Claude Code - the file does the production, the judgement stays yours ↓
1:13

ANTHROPIC LEAKED A FILE THEY SPENT 4 YEARS BUILDING - IT CREATES A MARKETING AGENCY IN 60 MINUTES AND GENERATES $100K A MONTH agency charged $3,000 a month not for the work but for 3 weeks of waiting - the actual work in there is 4 hours. here 6 roles read one brief at the same moment, so there's nobody to wait on. you write 4 lines at the start: the client, the tone, the offer, what we never do. the fourth line is the one an agency needs 3 meetings to establish. Opus 5 sets the positioning, Sonnet 5 runs 3 branches for posts, landing and visuals, Haiku 4.5 sorts where and when it all gets published. 3 variants exist before you finish reading the first - choosing what ships was always the job. 15 cents a day decide which hour every post lands on - the same post at the wrong hour is a post nobody saw. analytics is the only role that writes back into the brief - which angle died and where not to go again. cut that link and the other 5 hammer the same dead angle for a month. 6 roles at $2 a day against $3,000 a month - a 50x gap, and production stops being the bottleneck. save this and paste it into Claude Code - the file does the production, the judgement stays yours ↓

Sprytix

17,882 просмотров • 1 месяц назад

watch the small one on the left. it hears something at 00:05 and does nothing clever with it it just passes it on that is the whole trick. echo isolates a trace in ambient noise. moth follows the noise. kite finds a shorter route to it. vex starts collecting fragments. halo keeps the pattern so nobody has to look for it twice. nobody assigned any of that coherence in the top corner opens the cycle at 18%. by the time the wave crosses the field it is past 75%, and not one instruction was typed while it happened most people building agent teams do this backwards. they write a dispatcher, a queue, a state machine, and then wonder why five bots are slower than one. five bots stay slower than one until they can read each other's job descriptions and hand work sideways with no human standing in the middle three ways this breaks before it works: → give every bot the same description and the orchestrator has nothing to route on. every task comes straight back to you → let two bots own the same folder and they overwrite each other for six hours. you wake up to a clean, empty, perfectly synced nothing → skip the boundary line in one charter and that bot will send something. once is plenty to learn that lesson twenty agents run on one seat here. cheaper than a standup for three humans, and none of them ask what the priority is most setups have one agent and a bottleneck made of a person. this one has twenty, and the interesting part happens while nobody is watching full charters, the routing rule and the one line that stops a bot mid-send are in the piece below
0:30

watch the small one on the left. it hears something at 00:05 and does nothing clever with it it just passes it on that is the whole trick. echo isolates a trace in ambient noise. moth follows the noise. kite finds a shorter route to it. vex starts collecting fragments. halo keeps the pattern so nobody has to look for it twice. nobody assigned any of that coherence in the top corner opens the cycle at 18%. by the time the wave crosses the field it is past 75%, and not one instruction was typed while it happened most people building agent teams do this backwards. they write a dispatcher, a queue, a state machine, and then wonder why five bots are slower than one. five bots stay slower than one until they can read each other's job descriptions and hand work sideways with no human standing in the middle three ways this breaks before it works: → give every bot the same description and the orchestrator has nothing to route on. every task comes straight back to you → let two bots own the same folder and they overwrite each other for six hours. you wake up to a clean, empty, perfectly synced nothing → skip the boundary line in one charter and that bot will send something. once is plenty to learn that lesson twenty agents run on one seat here. cheaper than a standup for three humans, and none of them ask what the priority is most setups have one agent and a bottleneck made of a person. this one has twenty, and the interesting part happens while nobody is watching full charters, the routing rule and the one line that stops a bot mid-send are in the piece below

monokern

12,612 просмотров • 22 дней назад

GROK BOT + JEV ran 70 trades in one month with a $200 cap on each, and a $10,000 account ended at $20,000... two models build the desk, six agents run it, and one rule sits on every order: lose $200 at most, take $400 when it's there scan → vet → size → fill → hold → go Elon's Grok Bot finds the trade and fires the order Jev fires the stop before the entry exists, so no order lands without one between them sits a desk of six agents. scan finds the setup, vet throws most of them out, size puts $200 at risk and never more fills gets the order in, risk holds every stop, and chief says go or wait nobody on that floor can widen a stop the spider in the video crawls the whole desk, from the first order down to the last line of the honest part 70 trades in the month. 40 hit the target at +$400, 30 hit the stop at -$200 40 × $400 is $16,000, 30 × $200 is $6,000, and the $10,000 left over doubles the account that's 57% of trades right at two-to-one, and everything else stands on that 57% same rule and same 70 trades at 50% right is +$7,000. at 40% it's +$2,800. at 33% the month ends at -$200 a stop caps the loss on a normal day, and a gap jumps straight over it. ten stops in a row is -$2,000, and no rule here makes that impossible six agents, 70 trades, $200 at risk on each, and 100% in the one month it was right 57% of the time ↓
1:13

GROK BOT + JEV ran 70 trades in one month with a $200 cap on each, and a $10,000 account ended at $20,000... two models build the desk, six agents run it, and one rule sits on every order: lose $200 at most, take $400 when it's there scan → vet → size → fill → hold → go Elon's Grok Bot finds the trade and fires the order Jev fires the stop before the entry exists, so no order lands without one between them sits a desk of six agents. scan finds the setup, vet throws most of them out, size puts $200 at risk and never more fills gets the order in, risk holds every stop, and chief says go or wait nobody on that floor can widen a stop the spider in the video crawls the whole desk, from the first order down to the last line of the honest part 70 trades in the month. 40 hit the target at +$400, 30 hit the stop at -$200 40 × $400 is $16,000, 30 × $200 is $6,000, and the $10,000 left over doubles the account that's 57% of trades right at two-to-one, and everything else stands on that 57% same rule and same 70 trades at 50% right is +$7,000. at 40% it's +$2,800. at 33% the month ends at -$200 a stop caps the loss on a normal day, and a gap jumps straight over it. ten stops in a row is -$2,000, and no rule here makes that impossible six agents, 70 trades, $200 at risk on each, and 100% in the one month it was right 57% of the time ↓

Sprytix

15,227 просмотров • 7 дней назад

A basic VPS runs $20-40 a month at most providers. This GitHub README has been telling people how to get the same thing for $0 since 2015. > It's called free-for-dev. 132k stars, still maintained daily, 1600+ people adding new free tiers and killing dead ones every week. Not 14-day trials. The rule: if it's not free for at least a year, it doesn't make the list. Inside: → Oracle Cloud - 2 servers, 12GB RAM, free forever, no time limit → AWS Lambda - run code 1 million times a month before you pay a cent → Cloudflare - unlimited websites, DNS, SSL, and a global CDN, free → Google Cloud Run - 2 million requests a month on the house One rule they never break: if the security certificate isn't free too, it gets deleted from the list. Every side project starts with the same 20 minutes of Googling "is this actually free." This repo already did that search for you, and keeps redoing it every week. Bookmark it before your next project makes you do that search all over again.
1:44

A basic VPS runs $20-40 a month at most providers. This GitHub README has been telling people how to get the same thing for $0 since 2015. > It's called free-for-dev. 132k stars, still maintained daily, 1600+ people adding new free tiers and killing dead ones every week. Not 14-day trials. The rule: if it's not free for at least a year, it doesn't make the list. Inside: → Oracle Cloud - 2 servers, 12GB RAM, free forever, no time limit → AWS Lambda - run code 1 million times a month before you pay a cent → Cloudflare - unlimited websites, DNS, SSL, and a global CDN, free → Google Cloud Run - 2 million requests a month on the house One rule they never break: if the security certificate isn't free too, it gets deleted from the list. Every side project starts with the same 20 minutes of Googling "is this actually free." This repo already did that search for you, and keeps redoing it every week. Bookmark it before your next project makes you do that search all over again.

Atenov int.

182,777 просмотров • 1 месяц назад

My team and I have mentored 250+ founders in the last 6 months. Almost all of them were stuck on the same three things, and each one costs 6 to 12 months of progress (recorded a 17-minute breakdown of what they are and how to avoid them): 1. Most were spread too thin. 3 or 4 ideas at once. A bit of AI automation, some crypto, consulting on the side. You don't hit your first million that way. Pick one problem. Solve it a thousand times. 2. The ones who did pick one lane usually picked the wrong one. Lead generation is where businesses actually pay 4 to 5 figures a month. The tool stack that used to need $25K a year now runs on $300. But 80-90% of operators are using that edge to send more of the same copy that stopped working years ago. Volume is nothing but a race to the bottom. 3. And the ones on the right lane quit too early. "Six months to six figures" made them think they should already be rich, so they bailed in month 4. Year 1, only 15% of businesses are profitable. Year 2 and 3, another 40% get there. The operators who stay past month 4 hold all the case studies when the rest are just starting. Worth your time if you're building anything in the AI space now:
17:14

My team and I have mentored 250+ founders in the last 6 months. Almost all of them were stuck on the same three things, and each one costs 6 to 12 months of progress (recorded a 17-minute breakdown of what they are and how to avoid them): 1. Most were spread too thin. 3 or 4 ideas at once. A bit of AI automation, some crypto, consulting on the side. You don't hit your first million that way. Pick one problem. Solve it a thousand times. 2. The ones who did pick one lane usually picked the wrong one. Lead generation is where businesses actually pay 4 to 5 figures a month. The tool stack that used to need $25K a year now runs on $300. But 80-90% of operators are using that edge to send more of the same copy that stopped working years ago. Volume is nothing but a race to the bottom. 3. And the ones on the right lane quit too early. "Six months to six figures" made them think they should already be rich, so they bailed in month 4. Year 1, only 15% of businesses are profitable. Year 2 and 3, another 40% get there. The operators who stay past month 4 hold all the case studies when the rest are just starting. Worth your time if you're building anything in the AI space now:

Alex Vacca

10,507 просмотров • 6 месяцев назад

Persi Diaconis walked into a lecture at the University of Washington, held up a coin, and told a room of physicists that Richard Feynman was fooled by it his entire life. He was right. In 2007, Diaconis proved a coin flip is not 50/50. It lands on the side it started on about 51 percent of the time. The bias comes from the physics of rotation under gravity. Every physicist since Newton had assumed 50/50 without ever testing it. Every trading model built on that assumption is running on the same lie. The lecture was on Feynman's book "The Meaning of it All." Diaconis quoted the most famous line in it: "the first principle is that you must not fool yourself, and you are the easiest person to fool." Then he pointed out that Feynman himself was fooled by every coin he ever flipped. Feynman's own rule would have killed the 50/50 assumption on day one. The market is the same setup at scale. Every model that assumes independent 50/50 outcomes at the base layer is built on a physical impossibility. Order flow, positioning, forced flows, expiries all leave biases larger than 1 percent. Your gut cannot see them. The math already knows they are there. Diaconis's rule: before you trust a random process, check it. Actually check it. Not with a simulation. With a proof or an experiment. The coin is where you start. The chart is where the same rule pays. The only random thing about markets is how thoroughly you refuse to check them.
47:48

Persi Diaconis walked into a lecture at the University of Washington, held up a coin, and told a room of physicists that Richard Feynman was fooled by it his entire life. He was right. In 2007, Diaconis proved a coin flip is not 50/50. It lands on the side it started on about 51 percent of the time. The bias comes from the physics of rotation under gravity. Every physicist since Newton had assumed 50/50 without ever testing it. Every trading model built on that assumption is running on the same lie. The lecture was on Feynman's book "The Meaning of it All." Diaconis quoted the most famous line in it: "the first principle is that you must not fool yourself, and you are the easiest person to fool." Then he pointed out that Feynman himself was fooled by every coin he ever flipped. Feynman's own rule would have killed the 50/50 assumption on day one. The market is the same setup at scale. Every model that assumes independent 50/50 outcomes at the base layer is built on a physical impossibility. Order flow, positioning, forced flows, expiries all leave biases larger than 1 percent. Your gut cannot see them. The math already knows they are there. Diaconis's rule: before you trust a random process, check it. Actually check it. Not with a simulation. With a proof or an experiment. The coin is where you start. The chart is where the same rule pays. The only random thing about markets is how thoroughly you refuse to check them.

veles

18,798 просмотров • 2 месяцев назад

UC Berkeley just open-sourced FreeToken. (2–4x faster local LLM inference than Ollama) the results are wild: - Qwen3.6-35B on an 8GB GPU at 39.3 tokens/s - DeepSeek-V4-Flash 284B on a 32GB GPU at 22 tokens/s - GLM-5.2 753B on a 96GB GPU at 14.9 tokens/s a 35B model at 16-bit precision needs about 70GB just for its weights. even at 4 bits it is close to 18GB, and FreeToken serves it on an 8GB GPU. let me explain how: all three models mentioned above are Mixture-of-Experts, and that is what FreeToken takes advantage of. each layer holds hundreds of separate experts plus a small router that picks a few of them per token. Qwen3.6-35B activates roughly 3B of its 35B parameters per token. DeepSeek-V4-Flash picks 6 of 256 experts per layer, so 13B of its 284B run at a time. so compute was never the bottleneck. the weights a single step touches fit comfortably on a consumer GPU. every expert the router might pick still has to exist somewhere. they sit in system RAM, and the GPU keeps a cache of the ones the model has been using recently. so everything comes down to what happens when the router picks an expert that is not on the GPU. there are two ways to serve that miss: 1. copy it over PCIe and run it on the GPU 2. run it on the CPU, where it already lives both read from the same system memory, so they compete for one pool of bandwidth instead of adding to each other. existing engines pick one option and freeze it when the model loads. but routing changes on every token, so a fixed choice misses most of what the model asks for. FreeToken measures both bandwidths on your machine and splits each step's misses between the two paths in proportion. the GPU and CPU results then merge exactly, with no approximation. two machines with the same GPU can end up wanting opposite strategies, which I did not expect. a 5090 in a gaming desktop should push nearly everything over PCIe, while an 8GB laptop is better off computing most misses on the CPU. none of that is readable off a spec sheet, so the engine profiles it once per machine. the second half of the design is about agents. coding agents constantly rewrite their own history, and every edit normally forces thousands of tokens back through prefill. FreeToken saves its checkpoints at the exact boundaries agent frameworks cut on, so it only reprocesses the new part. its slowest first token stays under 44 seconds, while llama.cpp peaks at 232 and KTransformers at 946. it serves the OpenAI and Anthropic APIs under Apache 2.0, so Claude Code and Codex can point at it directly. releasing weights publicly decides who can download a model, not who can afford to run one. frontier open models keep shipping, and running them still assumes a rented cluster. meanwhile there are over a hundred million consumer machines with discrete GPUs sitting mostly idle. closing that gap was never a hardware problem, and work like this is what turns open weights into something you can actually use. paper: repo: almost every idea in this post, from why memory bandwidth decides the outcome to why moving weights costs more than computing on them, comes straight out of how a GPU is built. I wrote a detailed primer on that. the article is quoted below.
0:42

UC Berkeley just open-sourced FreeToken. (2–4x faster local LLM inference than Ollama) the results are wild: - Qwen3.6-35B on an 8GB GPU at 39.3 tokens/s - DeepSeek-V4-Flash 284B on a 32GB GPU at 22 tokens/s - GLM-5.2 753B on a 96GB GPU at 14.9 tokens/s a 35B model at 16-bit precision needs about 70GB just for its weights. even at 4 bits it is close to 18GB, and FreeToken serves it on an 8GB GPU. let me explain how: all three models mentioned above are Mixture-of-Experts, and that is what FreeToken takes advantage of. each layer holds hundreds of separate experts plus a small router that picks a few of them per token. Qwen3.6-35B activates roughly 3B of its 35B parameters per token. DeepSeek-V4-Flash picks 6 of 256 experts per layer, so 13B of its 284B run at a time. so compute was never the bottleneck. the weights a single step touches fit comfortably on a consumer GPU. every expert the router might pick still has to exist somewhere. they sit in system RAM, and the GPU keeps a cache of the ones the model has been using recently. so everything comes down to what happens when the router picks an expert that is not on the GPU. there are two ways to serve that miss: 1. copy it over PCIe and run it on the GPU 2. run it on the CPU, where it already lives both read from the same system memory, so they compete for one pool of bandwidth instead of adding to each other. existing engines pick one option and freeze it when the model loads. but routing changes on every token, so a fixed choice misses most of what the model asks for. FreeToken measures both bandwidths on your machine and splits each step's misses between the two paths in proportion. the GPU and CPU results then merge exactly, with no approximation. two machines with the same GPU can end up wanting opposite strategies, which I did not expect. a 5090 in a gaming desktop should push nearly everything over PCIe, while an 8GB laptop is better off computing most misses on the CPU. none of that is readable off a spec sheet, so the engine profiles it once per machine. the second half of the design is about agents. coding agents constantly rewrite their own history, and every edit normally forces thousands of tokens back through prefill. FreeToken saves its checkpoints at the exact boundaries agent frameworks cut on, so it only reprocesses the new part. its slowest first token stays under 44 seconds, while llama.cpp peaks at 232 and KTransformers at 946. it serves the OpenAI and Anthropic APIs under Apache 2.0, so Claude Code and Codex can point at it directly. releasing weights publicly decides who can download a model, not who can afford to run one. frontier open models keep shipping, and running them still assumes a rented cluster. meanwhile there are over a hundred million consumer machines with discrete GPUs sitting mostly idle. closing that gap was never a hardware problem, and work like this is what turns open weights into something you can actually use. paper: repo: almost every idea in this post, from why memory bandwidth decides the outcome to why moving weights costs more than computing on them, comes straight out of how a GPU is built. I wrote a detailed primer on that. the article is quoted below.

Akshay 🚀

344,809 просмотров • 1 месяц назад

EVERYONE BUYING A $4,200 MAC STUDIO FOR LOCAL AI MISSED THAT TESLA V100s FROM DGX SERVERS ARE DUMPED ON EBAY FOR $250, AND ONE JUST RAN GPT-OSS 20B AT HOME a Tesla V100 SXM2 16GB shipped inside nvidia DGX-1 nodes at $10,000 per card in 2017, eight per chassis. those same cards now retire from cloud datacenters as H100s land, dumped on ebay at $200 to $300. drop a $50 SXM2 to PCIe adapter into a Z8 G4, slot the card with thermal pad, total bill $330. GPT-OSS 20B runs at 87°C and 180W the article calls the $700 used RTX 3090 the best memory per dollar. this V100 hack does the same job for one third the price with HBM2. the catch: HP Z620 fails on AVX2, install CUDA 11 not 12, flip above 4G decoding in BIOS the V100 pulls 250W and the blower screams. the mac mini from the article pulls 30W and runs silent. $300 versus $4,200 buys a lot of electricity and earplugs over four years
1:24

EVERYONE BUYING A $4,200 MAC STUDIO FOR LOCAL AI MISSED THAT TESLA V100s FROM DGX SERVERS ARE DUMPED ON EBAY FOR $250, AND ONE JUST RAN GPT-OSS 20B AT HOME a Tesla V100 SXM2 16GB shipped inside nvidia DGX-1 nodes at $10,000 per card in 2017, eight per chassis. those same cards now retire from cloud datacenters as H100s land, dumped on ebay at $200 to $300. drop a $50 SXM2 to PCIe adapter into a Z8 G4, slot the card with thermal pad, total bill $330. GPT-OSS 20B runs at 87°C and 180W the article calls the $700 used RTX 3090 the best memory per dollar. this V100 hack does the same job for one third the price with HBM2. the catch: HP Z620 fails on AVX2, install CUDA 11 not 12, flip above 4G decoding in BIOS the V100 pulls 250W and the blower screams. the mac mini from the article pulls 30W and runs silent. $300 versus $4,200 buys a lot of electricity and earplugs over four years

Shadow Nick

103,257 просмотров • 3 месяцев назад

THE CHINESE INSTALLED 6 AI AGENTS RUNNING 24/7 FOR $11 A MONTH. THE BOX UNDER ONE DESK REPLACED $1,200 IN CLOUD SUBSCRIPTIONS. The Minisforum MS-S1 Max ships with 128GB of unified memory shared between CPU and GPU. Qwen3-Coder 30B runs at 40 to 50 tokens per second. 6 agents run simultaneously without anyone watching the meter. An inbox sorter drafts replies every 15 minutes. A research monitor watches 30+ feeds and lands a digest by 7 AM, a job that costs $15 to $20 a day on cloud APIs. The same stack on cloud subscriptions ran $800 to $1,200 a month. The hardware costs $3,000 once. Break-even hits month 3 to 4. The real shift is behavior, not savings. A meeting prep agent never would have justified per-token cost. Running it free overnight made it obvious.
0:53

THE CHINESE INSTALLED 6 AI AGENTS RUNNING 24/7 FOR $11 A MONTH. THE BOX UNDER ONE DESK REPLACED $1,200 IN CLOUD SUBSCRIPTIONS. The Minisforum MS-S1 Max ships with 128GB of unified memory shared between CPU and GPU. Qwen3-Coder 30B runs at 40 to 50 tokens per second. 6 agents run simultaneously without anyone watching the meter. An inbox sorter drafts replies every 15 minutes. A research monitor watches 30+ feeds and lands a digest by 7 AM, a job that costs $15 to $20 a day on cloud APIs. The same stack on cloud subscriptions ran $800 to $1,200 a month. The hardware costs $3,000 once. Break-even hits month 3 to 4. The real shift is behavior, not savings. A meeting prep agent never would have justified per-token cost. Running it free overnight made it obvious.

Lummox

14,093 просмотров • 3 месяцев назад

A broken laptop on the floor of your room turns into a server that bills $1,800 a month. Total spend: $55. The machine everyone throws away is the one that works for this. Dead battery, cracked bezel, boots in 4 minutes. None of that matters when nothing has to be fast. Get one: yours from the drawer, or a retired office laptop. Dell Latitude, HP EliteBook, ThinkPad T-series, 6th to 8th gen i5. Refurbishers list them at $80 to $150. Spend $55, no more: one stick of RAM to reach 16GB, one small SSD. RAM prices are climbing, so run stock until a job actually stalls. Install it in one evening: Ubuntu Server to a USB, boot the laptop from it, walk through setup. No desktop. Ollama in one line, then a 7B model quantized to 4-bit. 6 tokens a second. Too slow to demo live. Fast enough for work that runs while the room is dark. Service 1: invoice and receipt triage. $300 a month. Sell to bookkeepers, solo accountants, small trades. They forward a folder of receipts. The box extracts vendor, amount, date, category, appends every line to a ledger, and flags anything over $500. Every entry cites its source file. Easiest sale you will make, because one wrong entry costs them more than your fee. Service 2: email drafting. $500 a month. Sell to consultants, agencies, anyone drowning in a shared inbox. The box reads the last 5 messages in a thread, classifies routine versus needs-a-human, writes a reply in their voice, saves it to Drafts. Nothing sends. They open their inbox to a folder of finished replies and approve or delete. Highest perceived value, because they watch it work every single morning. Service 3: weekly digest. $400 a month. Sell to founders and consultants with messy note systems. Sunday night, the box reads everything touched that week, writes 3 to 5 updates, flags contradictions between new notes and old ones, keeps it under 400 words. Lowest effort on your side, set up once, bills monthly. Land the first client: offer 2 weeks free on invoice triage only. Do not pitch all 3. Hand back a filled ledger with source citations, then quote $300 for month one. Upsell drafting in month two. The math: 1 client on the bundle is $1,200 a month. 2 clients on the middle tier is $1,800. The laptop draws 20 watts, so electricity is $4 regardless. Margin sits above 99%, because inference costs nothing once the machine is yours. Where it caps out: around 4 clients. Past that, the box cannot finish overnight, and you upgrade with client money, never your own. It never stops, never asks for a raise, and never sends anything you did not read.
0:54

A broken laptop on the floor of your room turns into a server that bills $1,800 a month. Total spend: $55. The machine everyone throws away is the one that works for this. Dead battery, cracked bezel, boots in 4 minutes. None of that matters when nothing has to be fast. Get one: yours from the drawer, or a retired office laptop. Dell Latitude, HP EliteBook, ThinkPad T-series, 6th to 8th gen i5. Refurbishers list them at $80 to $150. Spend $55, no more: one stick of RAM to reach 16GB, one small SSD. RAM prices are climbing, so run stock until a job actually stalls. Install it in one evening: Ubuntu Server to a USB, boot the laptop from it, walk through setup. No desktop. Ollama in one line, then a 7B model quantized to 4-bit. 6 tokens a second. Too slow to demo live. Fast enough for work that runs while the room is dark. Service 1: invoice and receipt triage. $300 a month. Sell to bookkeepers, solo accountants, small trades. They forward a folder of receipts. The box extracts vendor, amount, date, category, appends every line to a ledger, and flags anything over $500. Every entry cites its source file. Easiest sale you will make, because one wrong entry costs them more than your fee. Service 2: email drafting. $500 a month. Sell to consultants, agencies, anyone drowning in a shared inbox. The box reads the last 5 messages in a thread, classifies routine versus needs-a-human, writes a reply in their voice, saves it to Drafts. Nothing sends. They open their inbox to a folder of finished replies and approve or delete. Highest perceived value, because they watch it work every single morning. Service 3: weekly digest. $400 a month. Sell to founders and consultants with messy note systems. Sunday night, the box reads everything touched that week, writes 3 to 5 updates, flags contradictions between new notes and old ones, keeps it under 400 words. Lowest effort on your side, set up once, bills monthly. Land the first client: offer 2 weeks free on invoice triage only. Do not pitch all 3. Hand back a filled ledger with source citations, then quote $300 for month one. Upsell drafting in month two. The math: 1 client on the bundle is $1,200 a month. 2 clients on the middle tier is $1,800. The laptop draws 20 watts, so electricity is $4 regardless. Margin sits above 99%, because inference costs nothing once the machine is yours. Where it caps out: around 4 clients. Past that, the box cannot finish overnight, and you upgrade with client money, never your own. It never stops, never asks for a raise, and never sends anything you did not read.

AiMind

10,307 просмотров • 2 месяцев назад

🚨 nasa you have to see this 50 Grok Bot agents rebuilt one of your own planets overnight for $200, and it looks better than anything in the archive TRAPPIST-1 e, catalogued in 2017, never once photographed: 0.92 Earth radii, 0.692 Earth masses, a year that lasts 6.101 days, 249 kelvin, orbiting 0.02925 AU from a 2566 kelvin red dwarf, 40 light years out, 7 worlds in the system. every figure burned into that panel came out of NASA's exoplanet archive over one public url and carries its own citation. the cities are mine and the frame says so turns out the build is five moves, and none of them are the ones people demo: 1. pull the data before a single frame exists. the archive answers a plain GET at takes sql, returns csv across all 6,000 confirmed planets. hand one bot that url and it stops inventing numbers forever 2. shard by shot, not by task. a bot drives one screen at a time, so speed comes from more bots and never from a smarter one. one bot per shot, all of them on their own cloud machines at once 3. keep every recorded skill under 10 minutes, because that is the hard cap. split it: "log in and start the render" is one skill, "check it and download it" is a second on a 5 minute schedule. a long job is a chain of short skills, not one long recording 4. do not let bots message each other to coordinate. they loop, and the loop quietly eats your quota. give them one shared folder and have each drop a file when its shot lands 5. run the last bot in a clean context holding the source csv and nothing else, and make it check every number on the panel against the row it came from honest ceiling: 10 minutes per recording and one screen per bot, so 50 of them buys width and not speed bookmark this. all 50 prompts go up in 24 hours ↓
0:29

🚨 nasa you have to see this 50 Grok Bot agents rebuilt one of your own planets overnight for $200, and it looks better than anything in the archive TRAPPIST-1 e, catalogued in 2017, never once photographed: 0.92 Earth radii, 0.692 Earth masses, a year that lasts 6.101 days, 249 kelvin, orbiting 0.02925 AU from a 2566 kelvin red dwarf, 40 light years out, 7 worlds in the system. every figure burned into that panel came out of NASA's exoplanet archive over one public url and carries its own citation. the cities are mine and the frame says so turns out the build is five moves, and none of them are the ones people demo: 1. pull the data before a single frame exists. the archive answers a plain GET at takes sql, returns csv across all 6,000 confirmed planets. hand one bot that url and it stops inventing numbers forever 2. shard by shot, not by task. a bot drives one screen at a time, so speed comes from more bots and never from a smarter one. one bot per shot, all of them on their own cloud machines at once 3. keep every recorded skill under 10 minutes, because that is the hard cap. split it: "log in and start the render" is one skill, "check it and download it" is a second on a 5 minute schedule. a long job is a chain of short skills, not one long recording 4. do not let bots message each other to coordinate. they loop, and the loop quietly eats your quota. give them one shared folder and have each drop a file when its shot lands 5. run the last bot in a clean context holding the source csv and nothing else, and make it check every number on the panel against the row it came from honest ceiling: 10 minutes per recording and one screen per bot, so 50 of them buys width and not speed bookmark this. all 50 prompts go up in 24 hours ↓

Argona

37,258 просмотров • 1 месяц назад

A woman just received her very first paycheck as a doctor and posted it online - and the comments are all over the place. What people do not realize is that a first year resident physician’s salary is actually one of the most humbling paychecks in professional America. You just spent 8 years in school, hundreds of thousands in student loans, and your first check as a doctor can run anywhere from $55,000 to $70,000 a year before taxes. But here is what the doubters are missing - that check does not stay there. Give it 5 to 10 years and that same doctor is pulling $300,000 to $500,000 or more depending on their specialty. The process is the process and every expert was once a beginner. She earned that check and every one after it is going to be bigger than the last.
1:23

A woman just received her very first paycheck as a doctor and posted it online - and the comments are all over the place. What people do not realize is that a first year resident physician’s salary is actually one of the most humbling paychecks in professional America. You just spent 8 years in school, hundreds of thousands in student loans, and your first check as a doctor can run anywhere from $55,000 to $70,000 a year before taxes. But here is what the doubters are missing - that check does not stay there. Give it 5 to 10 years and that same doctor is pulling $300,000 to $500,000 or more depending on their specialty. The process is the process and every expert was once a beginner. She earned that check and every one after it is going to be bigger than the last.

Tony Lane 🇺🇸

2,120,492 просмотров • 3 месяцев назад