正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

Packaging machine

Interesting AF

2,595,387 subscribers

14,023,357 次观看 • 1 年前 •via X (Twitter)

科学技术教育艺术

Anya Rossi• Live Now

Private livecam show

10 条评论

🐺FcLrIyGpHtTic🚀 的头像

🐺FcLrIyGpHtTic🚀1 年前

First of al!! I'm am not putting my arms in there 😌😐

Trending blog room 的头像

Trending blog room1 年前

But he’s still learning

KH 的头像

KH1 年前

In 1978 I worked at the local newspaper and I ran one like that to bundle the papers. Well, it was bigger and clunkier but did the same thing.

Gerson 的头像

Gerson1 年前

That’s not a packaging machine It’s a strapping machine

Trending blog room 的头像

Trending blog room1 年前

It’s very nice

OWONGO ON YOUTUBE 的头像

OWONGO ON YOUTUBE1 年前

Easy

Trending blog room 的头像

Trending blog room1 年前

And a smart technology

LastLaf 的头像

LastLaf1 年前

Operator needs to be put through head first, several times!🛴No?

恥丘人的头像

恥丘人1 年前

There is something to be learned from watching someone who is not good at it.

Michael M 的头像

Michael M1 年前

Can someone stop him packaging all that chinese shit?

相关视频

packaging machine testing

packaging machine testing

𝕾𝖒𝖎𝖑𝖊

425,698 次观看 • 7 个月前

packaging machine testing

packaging machine testing

Kaleeeee

1,295,435 次观看 • 7 个月前

If you thought the Slim style pre-roll machine was cool🙏 Check out the automated packaging machine we use for our 10 packs💪 this is the only one in Canada and one of three in the world for cannabis. National program starts now 🔌

If you thought the Slim style pre-roll machine was cool🙏 Check out the automated packaging machine we use for our 10 packs💪 this is the only one in Canada and one of three in the world for cannabis. National program starts now 🔌

Chris C

31,969 次观看 • 3 年前

A humanoid robot was shown packing clothes using an automatic machine by the company Robros. It picks up folded garments and places them into the machine for sealing and packaging. The robot is remotely controlled by a human, showing how robots can assist in factory packing tasks.

A humanoid robot was shown packing clothes using an automatic machine by the company Robros. It picks up folded garments and places them into the machine for sealing and packaging. The robot is remotely controlled by a human, showing how robots can assist in factory packing tasks.

Space and Technology

12,352 次观看 • 4 个月前

The Africa Boom — You're Not Buying a Machine, You're Building a Nation 💰Stop importing products. Start importing FACTORIES. From Toilet Paper Machines to Packaging Lines—how African founders are leveraging the Chinese ecosystem to dominate local mar...

The Africa Boom — You're Not Buying a Machine, You're Building a Nation 💰Stop importing products. Start importing FACTORIES. From Toilet Paper Machines to Packaging Lines—how African founders are leveraging the Chinese ecosystem to dominate local mar...

Eric Cracks China

18,327 次观看 • 3 个月前

There's a surprising amount of science behind the packaging your order arrives in. At Amazon's fulfillment centers, machine learning algorithms help analyze millions of products to figure out the best way to package each one. They determine what size, what material, and whether it even needs added packaging at all. The result: protecting products with less waste, one package at a time. Take a look at how it works. ⬇️

There's a surprising amount of science behind the packaging your order arrives in. At Amazon's fulfillment centers, machine learning algorithms help analyze millions of products to figure out the best way to package each one. They determine what size, what material, and whether it even needs added packaging at all. The result: protecting products with less waste, one package at a time. Take a look at how it works. ⬇️

Amazon

174,245 次观看 • 1 个月前

AI Engineers gonna love this MCP server! KitOps MCP serve eliminates the "it works on my machine" problem by packaging AI/ML projects (models, datasets, code, config) into a shareable, deployable ModelKit. Sharing projects is as easy as sharing zip files! 100% open-source.

AI Engineers gonna love this MCP server! KitOps MCP serve eliminates the "it works on my machine" problem by packaging AI/ML projects (models, datasets, code, config) into a shareable, deployable ModelKit. Sharing projects is as easy as sharing zip files! 100% open-source.

Akshay 🚀

87,067 次观看 • 1 年前

The FFXIV Eorzea Cafe building, called Pasela Resort in Akihabara, has a vending machine outside where they sell fizzy lemonade with a Scion sticker on the packaging! The design is random, so you won’t know who you get! Each drink is 800yen, and I got Krile and Urianger 🥹

The FFXIV Eorzea Cafe building, called Pasela Resort in Akihabara, has a vending machine outside where they sell fizzy lemonade with a Scion sticker on the packaging! The design is random, so you won’t know who you get! Each drink is 800yen, and I got Krile and Urianger 🥹

☆オードリーAudrey☆

69,348 次观看 • 4 个月前

How to pack products with class? 👀 Innovative product packaging by Douglas Machine! 📦 These robots can pack items either standing up or lying flat in the case, making it easy to create neat and appealing arrangements. They're designed to quickly switch between different packing styles, accommodating different products and layouts. Using robot arms, they handle everything from getting the products to loading them into the case, all in one machine. Thoughts? 👀 ~~ ♻️ Join the weekly robotics newsletter, and never miss any news →

How to pack products with class? 👀 Innovative product packaging by Douglas Machine! 📦 These robots can pack items either standing up or lying flat in the case, making it easy to create neat and appealing arrangements. They're designed to quickly switch between different packing styles, accommodating different products and layouts. Using robot arms, they handle everything from getting the products to loading them into the case, all in one machine. Thoughts? 👀 ~~ ♻️ Join the weekly robotics newsletter, and never miss any news →

Lukas Ziegler

13,627 次观看 • 5 个月前

Sebelum launching esok, i want to share with you guys journey kita develop Dendeng Nyet Berapi dari setahun lepas. Bermula dengan R&D resipi, pembelian mesin, product design, packaging material, factory renovation, machine test run, coaching, production kickoff, photoshoot, videoshoot, live setup dan sebagainya. Nak buat sesuatu produk memerlukan r&d yang sangat teliti untuk pastikan semua proses berjalan lancar. Thanks to my team, suppliers, vendors and everyone yang terlibat dalam produk kedua ni. We’ve got a big day tomorrow, jumpa you guys esok! 🥹

Sebelum launching esok, i want to share with you guys journey kita develop Dendeng Nyet Berapi dari setahun lepas. Bermula dengan R&D resipi, pembelian mesin, product design, packaging material, factory renovation, machine test run, coaching, production kickoff, photoshoot, videoshoot, live setup dan sebagainya. Nak buat sesuatu produk memerlukan r&d yang sangat teliti untuk pastikan semua proses berjalan lancar. Thanks to my team, suppliers, vendors and everyone yang terlibat dalam produk kedua ni. We’ve got a big day tomorrow, jumpa you guys esok! 🥹

Khairulaming

2,014,935 次观看 • 2 年前

Robotics is entering the mainstream! 👀 Let's accept this fact. Consumptionism – stonks. Possibilities (HW + SW) – stonks. Demographics - not stonks. Modern robots are often designed to interact with our legacy world, whether it's a household, a factory, or a shipyard. That means they often need to work with existing machines, tools, and spaces. In short: retrofitting. Improving old machines with new technologies without major interference in their operation. Modern way of integrating things. You don't have to change the entire environment to deploy a robot. The robot will adapt, just as human would when freshly hired. The promise is to have an army of robots to help with the world designed for us. I also believe that general purpose robots, are not a direct competition for single purpose machines (like a highly efficient specialized packaging machine, for example). It's like adding a robot to this carriage. Specialized machine could be a car ;) ~~ ♻️ Join the weekly robotics newsletter, and never miss any news →

Robotics is entering the mainstream! 👀 Let's accept this fact. Consumptionism – stonks. Possibilities (HW + SW) – stonks. Demographics - not stonks. Modern robots are often designed to interact with our legacy world, whether it's a household, a factory, or a shipyard. That means they often need to work with existing machines, tools, and spaces. In short: retrofitting. Improving old machines with new technologies without major interference in their operation. Modern way of integrating things. You don't have to change the entire environment to deploy a robot. The robot will adapt, just as human would when freshly hired. The promise is to have an army of robots to help with the world designed for us. I also believe that general purpose robots, are not a direct competition for single purpose machines (like a highly efficient specialized packaging machine, for example). It's like adding a robot to this carriage. Specialized machine could be a car ;) ~~ ♻️ Join the weekly robotics newsletter, and never miss any news →

Lukas Ziegler

14,396 次观看 • 6 个月前

Just look at these systems from CPS... they show what hyper-optimized automation actually looks like. Product arrives flat, a racetrack mechanism flips it upright, rows get grouped, and a bomb-door system drops the whole batch into the case. Fast, reliable, and dead simple. I don't see any serious humanoid adoption on the shop floor. Purpose-built machinery keep winning the speed and cost debate... at least for some more time to come. Packing flexible bags into cases looks easy until you try doing it at scale across different pouch formats. Seeing setups like this always brings up two questions for me. First, how does a flashy humanoid like Figure actually compare to a dedicated machine like this in the real world? Humanoids get all the media attention, but when it comes to raw throughput, uptime, and ROI on a specific packaging line, specialized kinematics usually wipe the floor with general-purpose robots. Second, how long until European regulations try to over-engineer safety or compliance rules around setups like this? While engineers are trying to innovate and speed up factory floors, we seem remarkably good at inventing new regulatory hurdles that slow down deployment across Europe. I’m curious how people who are working in those startups, packaging and industrial automation view this situation.

Just look at these systems from CPS... they show what hyper-optimized automation actually looks like. Product arrives flat, a racetrack mechanism flips it upright, rows get grouped, and a bomb-door system drops the whole batch into the case. Fast, reliable, and dead simple. I don't see any serious humanoid adoption on the shop floor. Purpose-built machinery keep winning the speed and cost debate... at least for some more time to come. Packing flexible bags into cases looks easy until you try doing it at scale across different pouch formats. Seeing setups like this always brings up two questions for me. First, how does a flashy humanoid like Figure actually compare to a dedicated machine like this in the real world? Humanoids get all the media attention, but when it comes to raw throughput, uptime, and ROI on a specific packaging line, specialized kinematics usually wipe the floor with general-purpose robots. Second, how long until European regulations try to over-engineer safety or compliance rules around setups like this? While engineers are trying to innovate and speed up factory floors, we seem remarkably good at inventing new regulatory hurdles that slow down deployment across Europe. I’m curious how people who are working in those startups, packaging and industrial automation view this situation.

Ilir Aliu

56,157 次观看 • 12 天前

When Did People Magazine Become Meghan Markle’s Spokesperson? People Magazine has published an astonishing volume of Meghan Markle articles just in 2026 alone. Not major interviews. Not award wins. Not blockbuster projects. Not cultural moments that actually dominated headlines. Instead, readers get a relentless cycle of product launches, Instagram recaps, outfit breakdowns, “sweet family nods,” vacation glimpses, and glowing lifestyle features. Month after month, the same stories recycled in slightly different packaging. The formula never changes: Meghan Markle the devoted mother. Meghan the loving wife. Meghan the visionary entrepreneur. Meghan the misunderstood figure. Meghan the resilient success story. Different headline, identical narrative. After well over a hundred articles in 2026 alone built around the same themes and the same flattering framing, the question is whether they’re functioning as journalists or as her personal mouthpiece. If the goal is to churn out endless celebrity fluff pieces with no pushback, context, scrutiny, or balance, then call it what it is. A celebrity image-management platform disguised as reporting. People Magazine has devolved into a captive mouthpiece for Meghan Markle’s image rehabilitation machine. The archive speaks for itself. P.S. After compiling 100+ Meghan Markle headlines from People Magazine, I got tired. There are still more to go. 🥱

When Did People Magazine Become Meghan Markle’s Spokesperson? People Magazine has published an astonishing volume of Meghan Markle articles just in 2026 alone. Not major interviews. Not award wins. Not blockbuster projects. Not cultural moments that actually dominated headlines. Instead, readers get a relentless cycle of product launches, Instagram recaps, outfit breakdowns, “sweet family nods,” vacation glimpses, and glowing lifestyle features. Month after month, the same stories recycled in slightly different packaging. The formula never changes: Meghan Markle the devoted mother. Meghan the loving wife. Meghan the visionary entrepreneur. Meghan the misunderstood figure. Meghan the resilient success story. Different headline, identical narrative. After well over a hundred articles in 2026 alone built around the same themes and the same flattering framing, the question is whether they’re functioning as journalists or as her personal mouthpiece. If the goal is to churn out endless celebrity fluff pieces with no pushback, context, scrutiny, or balance, then call it what it is. A celebrity image-management platform disguised as reporting. People Magazine has devolved into a captive mouthpiece for Meghan Markle’s image rehabilitation machine. The archive speaks for itself. P.S. After compiling 100+ Meghan Markle headlines from People Magazine, I got tired. There are still more to go. 🥱

Queen Esther

53,675 次观看 • 3 个月前

🧵🔥👀 WOW… YOU CANT UNSEE IT. This is a LONG thread but it’s all right there plain as day. THE OBAMA MACHINE HAS BEEN BEHIND GETTING MAMDANI ELECTED & SOON HAVING GRAHAM PLATNER (WHO ACCIDENTALLY GOT A NAZI TATTOO HE COVERED UP) INSTALLED AS SENATOR. So Who is MORRIS KATZ? What’s he have to do with Hollywood & Spielberg? I’ve been saying this for OVER A YEAR… Platner is gonna be the Democrat nominee and hes gonna defeat RINO Collins. It’s The same forces / PR Machines and ACTORS used to go out and select the same people as AOC & MAMDANI. What if I told you David Axelrod who is Obama’s boy and former Campaign manager was in constant communication with Morris Katz who is an advisor for Mandani. Katz decided to work to push him in early 2024 when Mamdani was only polling at 1%. What if I told you he they went out and SELECTED Platner last year? Now what I told you that when Platners nice packaging fell apart because he had a TOTTENKAMP (THE S.S. DEATH’S HEAD TATTOO ON HIS CHEST) that Graham appeared on POD SAVE AMERICA for damage control first? What if I told you that Jon Favreau the Obama speechwriter and Dan Pfeiffer host that podcast that were selected as assets of the PODCAST machine because they were former OBAMA SENIORS AIDES. President Trump should have been running a serious primary candidate instead of Collins and primaried her but ONCE AGAIN he’s being given bad advice. Why do all 3 of Morris Katz client candidates have almost SIMILAR AW SHUCKS AMERICANA IM JUST LIKE YOU HERE IN (INSERT CITY/STATE) AD CAMPAIGNS & ANTI BILLIONAIRE MESSAGING. Why did the mockingbird media not obsess over a Democrat that had this tattoo… yet endlessly vilified Elon for a gesture and Hegseth for his CHRISTIAN TATTOOS. Remember how Swalwell was elevated … placed on Intel committes, Jan 6 committee but all along he was a COMPROMISED FOOL… who does Graham Platner remind you of?? 👇🏻

🧵🔥👀 WOW… YOU CANT UNSEE IT. This is a LONG thread but it’s all right there plain as day. THE OBAMA MACHINE HAS BEEN BEHIND GETTING MAMDANI ELECTED & SOON HAVING GRAHAM PLATNER (WHO ACCIDENTALLY GOT A NAZI TATTOO HE COVERED UP) INSTALLED AS SENATOR. So Who is MORRIS KATZ? What’s he have to do with Hollywood & Spielberg? I’ve been saying this for OVER A YEAR… Platner is gonna be the Democrat nominee and hes gonna defeat RINO Collins. It’s The same forces / PR Machines and ACTORS used to go out and select the same people as AOC & MAMDANI. What if I told you David Axelrod who is Obama’s boy and former Campaign manager was in constant communication with Morris Katz who is an advisor for Mandani. Katz decided to work to push him in early 2024 when Mamdani was only polling at 1%. What if I told you he they went out and SELECTED Platner last year? Now what I told you that when Platners nice packaging fell apart because he had a TOTTENKAMP (THE S.S. DEATH’S HEAD TATTOO ON HIS CHEST) that Graham appeared on POD SAVE AMERICA for damage control first? What if I told you that Jon Favreau the Obama speechwriter and Dan Pfeiffer host that podcast that were selected as assets of the PODCAST machine because they were former OBAMA SENIORS AIDES. President Trump should have been running a serious primary candidate instead of Collins and primaried her but ONCE AGAIN he’s being given bad advice. Why do all 3 of Morris Katz client candidates have almost SIMILAR AW SHUCKS AMERICANA IM JUST LIKE YOU HERE IN (INSERT CITY/STATE) AD CAMPAIGNS & ANTI BILLIONAIRE MESSAGING. Why did the mockingbird media not obsess over a Democrat that had this tattoo… yet endlessly vilified Elon for a gesture and Hegseth for his CHRISTIAN TATTOOS. Remember how Swalwell was elevated … placed on Intel committes, Jan 6 committee but all along he was a COMPROMISED FOOL… who does Graham Platner remind you of?? 👇🏻

Johnny St.Pete

14,247 次观看 • 4 个月前

UN’S ARCHITECT OF OPEN BORDERS Meet Amy Pope — the most powerful globalist you’ve never heard of. The UN’s migration chief, working hand-in-hand with Pope Leo XIV to erase borders and dismantle Christian civilization under the banner of “compassion.” 🔹 WHO SHE IS – Former Obama–Biden insider, Deputy Homeland Security Advisor under Obama – Architect of the “Countering Violent Extremism” program that erased Islam and jihad from FBI training – Creator of Obama’s #RefugeesWelcome campaign — launched days after the Paris terror attacks to shame Americans into silence 🔹 WHAT SHE RUNS – Director General of the UN’s International Organization for Migration (IOM) – 17,000-person bureaucracy, $2.5 billion budget, 67 % funded by U.S. taxpayers – Moves populations across the West while calling it “humanitarian aid” 🔹 WHO SHE WORKS WITH – Pope Leo XIV, who calls mass migration a “new missionary age” — turning border collapse into moral virtue – GWL Voices, Davos, and NGO networks packaging globalism as “progress” – Tax-funded “religious charities” seeding UN-selected refugees into U.S. towns without consent This isn’t mercy. It’s the management of the West’s decline — coordinated by the UN and sanctified by Rome. This is how we end it: Defund the IOM. Cut off the money fueling the UN’s migration machine. Withdraw from the UN’s migration network. Take back control of our borders and policies. Reclaim the Christian West. Protect the faith, the nations, and the civilization we were given to defend. See my complete exposé here:

UN’S ARCHITECT OF OPEN BORDERS Meet Amy Pope — the most powerful globalist you’ve never heard of. The UN’s migration chief, working hand-in-hand with Pope Leo XIV to erase borders and dismantle Christian civilization under the banner of “compassion.” 🔹 WHO SHE IS – Former Obama–Biden insider, Deputy Homeland Security Advisor under Obama – Architect of the “Countering Violent Extremism” program that erased Islam and jihad from FBI training – Creator of Obama’s #RefugeesWelcome campaign — launched days after the Paris terror attacks to shame Americans into silence 🔹 WHAT SHE RUNS – Director General of the UN’s International Organization for Migration (IOM) – 17,000-person bureaucracy, $2.5 billion budget, 67 % funded by U.S. taxpayers – Moves populations across the West while calling it “humanitarian aid” 🔹 WHO SHE WORKS WITH – Pope Leo XIV, who calls mass migration a “new missionary age” — turning border collapse into moral virtue – GWL Voices, Davos, and NGO networks packaging globalism as “progress” – Tax-funded “religious charities” seeding UN-selected refugees into U.S. towns without consent This isn’t mercy. It’s the management of the West’s decline — coordinated by the UN and sanctified by Rome. This is how we end it: Defund the IOM. Cut off the money fueling the UN’s migration machine. Withdraw from the UN’s migration network. Take back control of our borders and policies. Reclaim the Christian West. Protect the faith, the nations, and the civilization we were given to defend. See my complete exposé here:

Amy Mek

3,943,642 次观看 • 11 个月前

ACCELERATOR-BASED LITHOGRAPHY AND THE INDUCTION STORAGE RING LIGHT SOURCE The future of chipmaking may look less like a factory and more like a power grid. The proposed “Terafab” paradigm reimagines semiconductor manufacturing at utility scale: chip design, wafer fabrication, EUV lithography, memory, advanced packaging, and testing all under one roof, with the ambition of producing more than one terawatt of AI compute capacity per year. Its most radical innovation is treating light as a utility. Today, each EUV scanner relies on its own laser-produced plasma source, firing lasers at molten tin to generate 13.5 nm light. The process is inefficient, intensely hot, debris-heavy, and difficult to scale. Terafab replaces those individual sources with centralized, accelerator-driven free-electron lasers capable of distributing multi-kilowatt EUV light across an entire network of scanners. The advantages could be transformative: • No tin contamination or destructive plasma debris • Higher efficiency through electron-beam energy recovery • Greater photon flux to suppress stochastic defects at sub-3 nm nodes • Redundant accelerators that keep scanners operating during maintenance • Tunable wavelengths, potentially enabling 6.x nm “Beyond EUV” lithography THAT LAST POINT MATTERS ENORMOUSLY Conventional EUV is locked to the atomic emission of tin. Free-electron lasers are not. Their wavelength can be tuned through the energy of the electron beam creating a possible path beyond today’s 13.5 nm limit. If realized, Terafab would represent more than a larger semiconductor plant. It would transform lithography from a collection of isolated tools into shared industrial infrastructure. The next era of chipmaking may not be defined by a better machine. It may be defined by an entirely new architecture for manufacturing intelligence at civilization scale.

ACCELERATOR-BASED LITHOGRAPHY AND THE INDUCTION STORAGE RING LIGHT SOURCE The future of chipmaking may look less like a factory and more like a power grid. The proposed “Terafab” paradigm reimagines semiconductor manufacturing at utility scale: chip design, wafer fabrication, EUV lithography, memory, advanced packaging, and testing all under one roof, with the ambition of producing more than one terawatt of AI compute capacity per year. Its most radical innovation is treating light as a utility. Today, each EUV scanner relies on its own laser-produced plasma source, firing lasers at molten tin to generate 13.5 nm light. The process is inefficient, intensely hot, debris-heavy, and difficult to scale. Terafab replaces those individual sources with centralized, accelerator-driven free-electron lasers capable of distributing multi-kilowatt EUV light across an entire network of scanners. The advantages could be transformative: • No tin contamination or destructive plasma debris • Higher efficiency through electron-beam energy recovery • Greater photon flux to suppress stochastic defects at sub-3 nm nodes • Redundant accelerators that keep scanners operating during maintenance • Tunable wavelengths, potentially enabling 6.x nm “Beyond EUV” lithography THAT LAST POINT MATTERS ENORMOUSLY Conventional EUV is locked to the atomic emission of tin. Free-electron lasers are not. Their wavelength can be tuned through the energy of the electron beam creating a possible path beyond today’s 13.5 nm limit. If realized, Terafab would represent more than a larger semiconductor plant. It would transform lithography from a collection of isolated tools into shared industrial infrastructure. The next era of chipmaking may not be defined by a better machine. It may be defined by an entirely new architecture for manufacturing intelligence at civilization scale.

Lacey

12,579 次观看 • 26 天前

This Chinese mathematician earned $10,000 a month inventing the hardest problems to train Neural Networks through Scale AI. Today his income dropped to zero. All the solutions are now generated by the model itself. He used to just hold the problem in his head and spell it out in plain text. His work is pure intellect. An expert in higher mathematics, he made his money hand-crafting the trickiest puzzles to test and train neural networks via RLHF. The bastion of "human" logic rested entirely on him, on people with PhDs who knew how to invent the problem. The collapse is simple. The shift to RLAIF and synthetic data. The model plays against itself, builds trees of logical inference, and solves deeper than a human can even invent the problem. No PhD data engineers, no hand-written prompt-completion examples, no manual grading. Just the model, search algorithms, and Chain of Thought. Ready-made "smart human-time" still sells on the market for many times more. His old rate was $50–100 per problem. The internal "mini-app" was written by the model too. Inside there's no pretty shell, just bare logic with exact steps: input: the problem statement inference tree: thousands of branches per second check: every step verifies itself output: a proof a human never had time to invent And here is what the whole setup looked like. He no longer needs to write an example by hand. He gave the model a direct instruction in human words, without a single formal term: "solve the problem yourself and grade yourself yourself" That's it. After that the algorithm found the solution, checked it, and trained on its own result, with no human. → the contractor got $50–100 per problem written → from 5,000 to 10,000 a month → now that income is annulled → a query to a math LLM costs 1–5 cents → a quant or an actuary runs 150,000–250,000 a year → the margin for whoever packages this into an agent is nearly 100% In the author's own words: "I'm no longer able to invent a problem the machine can't solve. The examiner became dumber than the one he's examining." But honestly, he admits the crude mistake himself, and it's not in the math, it's in the positioning. He tied his income to selling "smart human-time", to crafting formulas by hand. As long as he sells formulas, he's left behind. The machine computes faster than he can invent the problem. He names the right move himself: the role shifts from "intellectual craftsman" to "systems architect." Then he doesn't sell his time, he manages compute, packaging that same LLM into an autonomous agent that runs 24/7. Out of everything I've seen this year about the disappearance of intellectual professions, this is the most honest example: $50 per problem zeroed out to 1 cent per query, a doctor of science losing to a search algorithm, one problem stated in human words instead of a hand-written dataset, and right away an out-loud admission of the wrong business model. The barrier to entry in higher mathematics just dropped to the level of "describe the task in words." The only question is who'll be the first to stop selling their time and start managing the machine's compute.

This Chinese mathematician earned $10,000 a month inventing the hardest problems to train Neural Networks through Scale AI. Today his income dropped to zero. All the solutions are now generated by the model itself. He used to just hold the problem in his head and spell it out in plain text. His work is pure intellect. An expert in higher mathematics, he made his money hand-crafting the trickiest puzzles to test and train neural networks via RLHF. The bastion of "human" logic rested entirely on him, on people with PhDs who knew how to invent the problem. The collapse is simple. The shift to RLAIF and synthetic data. The model plays against itself, builds trees of logical inference, and solves deeper than a human can even invent the problem. No PhD data engineers, no hand-written prompt-completion examples, no manual grading. Just the model, search algorithms, and Chain of Thought. Ready-made "smart human-time" still sells on the market for many times more. His old rate was $50–100 per problem. The internal "mini-app" was written by the model too. Inside there's no pretty shell, just bare logic with exact steps: input: the problem statement inference tree: thousands of branches per second check: every step verifies itself output: a proof a human never had time to invent And here is what the whole setup looked like. He no longer needs to write an example by hand. He gave the model a direct instruction in human words, without a single formal term: "solve the problem yourself and grade yourself yourself" That's it. After that the algorithm found the solution, checked it, and trained on its own result, with no human. → the contractor got $50–100 per problem written → from 5,000 to 10,000 a month → now that income is annulled → a query to a math LLM costs 1–5 cents → a quant or an actuary runs 150,000–250,000 a year → the margin for whoever packages this into an agent is nearly 100% In the author's own words: "I'm no longer able to invent a problem the machine can't solve. The examiner became dumber than the one he's examining." But honestly, he admits the crude mistake himself, and it's not in the math, it's in the positioning. He tied his income to selling "smart human-time", to crafting formulas by hand. As long as he sells formulas, he's left behind. The machine computes faster than he can invent the problem. He names the right move himself: the role shifts from "intellectual craftsman" to "systems architect." Then he doesn't sell his time, he manages compute, packaging that same LLM into an autonomous agent that runs 24/7. Out of everything I've seen this year about the disappearance of intellectual professions, this is the most honest example: $50 per problem zeroed out to 1 cent per query, a doctor of science losing to a search algorithm, one problem stated in human words instead of a hand-written dataset, and right away an out-loud admission of the wrong business model. The barrier to entry in higher mathematics just dropped to the level of "describe the task in words." The only question is who'll be the first to stop selling their time and start managing the machine's compute.

Blaze

49,109 次观看 • 2 个月前

I've migrated the old Mast3r-SLAM example I had made last year to the latest version of Rerun and made a bunch of improvements! I wanted to spend some time with agents to modernize it. Here's an example of me walking around with my iPhone and getting a dense reconstruction at about 10FPS on a 5090. Heres the following improvements I made. Brought it into the monorepo with proper packaging: • Using prefix.dev pixi-build to get rid of all the mast3r/asmk/lietorch vendored code with just a few small patches. This let me remove so 60k lines of code from the repo! • Don't have to build the lietorch code on my machine anymore, which was taking ~10 minutes to compile (and also made it work on blackwell when it previously did not) Rebuilt the Gradio interface: • Fixed incremental updates, .MOV uploads, and stop behavior • Made the CLI + Gradio interface share the same entry point so updates automatically propagate Upgraded the Rerun integration: • Switched to a multiprocessing async logging strategy • Added video/pointmap/confidence logging • Improved blueprint layout and hid noisy entities from 3D view • Biggest perf win was the async background logger - documented about a ~2.5x speedup from decoupling logging from tracking The newest and most interesting part was my attempt to replace the CUDA kernels for Gauss-Newton ray matching with a Modular Mojo backend. As a Python dev, every time I look at CUDA code I basically shy away as it's pretty difficult for me to understand. Mojo let me rewrite the matching logic in a syntax I'm more comfortable with while still getting near-CUDA performance. Mojo is now the default matching backend with CUDA fallback. One major piece that's missing is the custom PyTorch op path, but I'll eventually do that as well. I heavily leaned on Claude Code to do the CUDA → Mojo migration, and I have no doubt it's not the cleanest or most idiomatic, BUT it's way more readable for me and helps me better understand the underlying algorithm. This was a ton of work, and a large part of why I'm doing it is how the monorepo compounds. This becomes an artifact for the next example I want to build with Claude that I can point to, which will make it even faster to implement. The compounding nature of this is really interesting and part of why I'm spending so much time trying to make things nice and readable.

I've migrated the old Mast3r-SLAM example I had made last year to the latest version of Rerun and made a bunch of improvements! I wanted to spend some time with agents to modernize it. Here's an example of me walking around with my iPhone and getting a dense reconstruction at about 10FPS on a 5090. Heres the following improvements I made. Brought it into the monorepo with proper packaging: • Using prefix.dev pixi-build to get rid of all the mast3r/asmk/lietorch vendored code with just a few small patches. This let me remove so 60k lines of code from the repo! • Don't have to build the lietorch code on my machine anymore, which was taking ~10 minutes to compile (and also made it work on blackwell when it previously did not) Rebuilt the Gradio interface: • Fixed incremental updates, .MOV uploads, and stop behavior • Made the CLI + Gradio interface share the same entry point so updates automatically propagate Upgraded the Rerun integration: • Switched to a multiprocessing async logging strategy • Added video/pointmap/confidence logging • Improved blueprint layout and hid noisy entities from 3D view • Biggest perf win was the async background logger - documented about a ~2.5x speedup from decoupling logging from tracking The newest and most interesting part was my attempt to replace the CUDA kernels for Gauss-Newton ray matching with a Modular Mojo backend. As a Python dev, every time I look at CUDA code I basically shy away as it's pretty difficult for me to understand. Mojo let me rewrite the matching logic in a syntax I'm more comfortable with while still getting near-CUDA performance. Mojo is now the default matching backend with CUDA fallback. One major piece that's missing is the custom PyTorch op path, but I'll eventually do that as well. I heavily leaned on Claude Code to do the CUDA → Mojo migration, and I have no doubt it's not the cleanest or most idiomatic, BUT it's way more readable for me and helps me better understand the underlying algorithm. This was a ton of work, and a large part of why I'm doing it is how the monorepo compounds. This becomes an artifact for the next example I want to build with Claude that I can point to, which will make it even faster to implement. The compounding nature of this is really interesting and part of why I'm spending so much time trying to make things nice and readable.

Pablo Vela

42,571 次观看 • 4 个月前