While Pyotr Kropotkin’s economic model in The Conquest of Bread was written as a theoretical vision, several historical revolutionary movements have attempted to put its specific principles—such as the immediate abolition of money, voluntary federation of agriculture and industry, communal distribution based on need, and bottom-up workers' assemblies—into direct practice.

The two most prominent attempts occurred during major 20th-century revolutions, along with smaller-scale autonomous implementations.

Major Practical Implementations

1. The Spanish Revolution (1936–1939)

The most comprehensive historical attempt to implement Kropotkin’s economic ideas took place in Spain during the Civil War, led by the anarcho-syndicalist union CNT (Confederación Nacional del Trabajo) and the FAI.

• Agricultural Collectives (Aragon, Levante, & Catalonia): In rural regions like Aragon, over 400 agricultural collectives were established. Peasants pooled land, draft animals, and machinery into common property. In direct alignment with Kropotkin's "Bread First" principle, villages focused on local food security before exporting surpluses to urban areas.

• Abolition of Money & Free Distribution: Many agrarian collectives abolished Spanish pesetas. Instead, they instituted family ration cards, local labor vouchers, or entirely free access to municipal communal stores ("free stores"), embodying Kropotkin's motto "To each according to their needs". Basic necessities—rent, healthcare, electricity, and education—were provided free of charge.

• Industrial Self-Management: In Barcelona and surrounding industrial centers, workers expropriated public transit, textile mills, utilities, and factories, reorganizing production through bottom-up factory committees (comités de empresa) rather than state administrators.

2. The Free Territory of Ukraine / Makhnovshchina (1918–1921)

During the Russian Civil War, the Revolutionary Insurgent Army of Ukraine, led by Nestor Makhno, established an autonomous territory covering parts of modern Southeastern Ukraine. Makhno and his comrades were explicitly influenced by Kropotkin’s works.

• Free Communes: Peasant communities organized agrarian communes (such as the Rosa Luxemburg Commune near Huliaipole) where land, tools, and livestock were held in common.

• Worker-Controlled Distribution: Peasants organized direct grain-for-manufactured-goods swaps with urban worker syndicates in cities like Moscow and Petrograd, bypassing state procurement agencies and middle-man markets.

• Free Assemblies: Regional congresses of workers, peasants, and insurgents met to coordinate defense and regional resource logistics without a central government.

Notable Precedents & Smaller Movements

Event / Region	Period	Key Practices Mirrored from The Conquest of Bread
The Paris Commune	1871	Though preceding the publication of The Conquest of Bread, Kropotkin drew heavy inspiration from it. Workers seized abandoned workshops, set up direct municipal distribution, and instituted democratic public service boards.
Strandzha Commune (Bulgarian Preobrazhenie Uprising)	1903	During a brief anti-Ottoman revolt, Bulgarian and Macedonian anarchists declared a revolutionary commune in the Strandzha region. They abolished taxation, expropriated food stockpiles, and established free collective distribution.
Shinmin Autonomous Region (Manchuria)	1929–1931	Organized by the Korean Anarchist Federation in Manchuria (CPM), over two million people established a decentralized network of cooperatives, free education systems, and self-managed agricultural communes before being overrun by Japanese military forces.

Primary Challenges Faced in Practice

In almost every case where Kropotkin’s model was attempted, the experiments were cut short not by internal economic collapse, but by external geopolitical and military forces:

1. Military Repression: Autonomous zones were violently suppressed by external enemies or competing revolutionary factions (e.g., Bolshevik forces crushing the Ukrainian Free Territory in 1921; Stalinist-backed Spanish Republican units and Francoist troops dismantling the Spanish collectives in 1937–1939).

2. War Economies: Running a decentralized, need-based economic system while simultaneously managing a total war effort against organized armies created massive resource shortages and logistical bottlenecks.

 The sixth mass extinction, formally known as the Holocene extinction or Anthropocene extinction, is an ongoing biodiversity crisis characterized by the rapid and widespread loss of Earth's plant and animal species. [1, 2]

A "mass extinction" is scientifically defined as an event where the planet loses roughly 75% of its total species within a short geological period. While Earth has experienced five previous mass extinctions over the last 500 million years—such as the asteroid that wiped out the dinosaurs 66 million years ago—the sixth mass extinction is entirely unique because it is driven directly by human activity rather than natural catastrophes. [1, 2, 3, 4]

To understand how current trends fit into Earth's history, you can watch a visual overview of past events and our current path from the American Museum of Natural History:

8:19

Six Extinctions, Viewed Through the Museum's Collections

111k views · 10 years ago

American Museum of Natural History · American Museum of Natural History

---

1. How Fast Is It Happening?

The core alarm raised by scientists is not just the total number of species already gone, but the speed of the collapse. [1]

• The Background Rate: Under normal, historical conditions, species naturally go extinct at a baseline rate of roughly 1 species per 10,000 species every 100 years. [1]
• The Current Rate: Modern extinction rates are estimated to be 100 to 1,000 times higher (and by some metrics up to 10,000 times higher) than the natural baseline. This means processes that normally take hundreds of thousands of years are unfolding in mere decades. [1, 2, 3]
• According to reports synthesized by the World Wildlife Fund (WWF), wildlife population abundance globally has plunged by nearly 70% over the last 50 years alone. [1]

2. The Primary Human Drivers

Scientists often point to a combination of human-induced pressures to explain the acceleration: [1]

Driver	Description	Impact
Habitat Loss & Fragmentation	Converting wilderness for cities, roads, and intensive agriculture.	Up to 40% of Earth's land has been converted for agriculture, destroying the primary homes of wildlife.
Overexploitation	Unsustainable industrial-scale hunting, poaching, and overfishing.	Collapses marine food webs and drives highly targeted vulnerable species toward extinction.
Climate Change	Burning fossil fuels to rapidly alter global temperatures, weather patterns, and ocean chemistry.	Forced displacement of ecosystems; ocean acidification severely impacts foundational species like coral reefs.
Invasive Species & Disease	Transporting non-native organisms around the globe via international trade.	Non-native species outcompete, prey upon, or introduce deadly pathogens to isolated local ecosystems.
Pollution	Discharging plastics, heavy metals, agricultural fertilizers, and toxic chemicals into air and waterways.	Poisons marine and terrestrial biomes, leading to phenomenon like "Insectageddon" (mass insect die-offs).

3. Why It Matters to Humanity

The extinction crisis is not just about losing beautiful animals; it is about the unraveling of the ecosystem services that keep humans alive. All life on Earth exists in a finely interwoven web. When foundational species (like pollinators, soil microbes, or coral reefs) disappear, the entire system risks a domino-effect collapse. This threatens our own access to: [1, 2, 3, 4]

• Clean air and fresh water filtration
• Natural crop pollination and global food security
• Climate stabilization and natural coastal defense boundaries
• Source materials for modern medicine [1, 2]

Is it preventable?

Unlike an asteroid strike, the sixth mass extinction is entirely within our control to mitigate. International frameworks like the 30x30 initiative (aiming to protect 30% of the planet's land and water by 2030) alongside aggressive climate actions, rewilding projects, and shifting away from habitat-destructive consumption patterns are key levers scientists use to slow down the collapse. [1, 2, 3]

Would you like to explore specific animal groups that are currently suffering the steepest declines (such as amphibians or insects), or look into how conservation strategies like rewilding work?

 

TSMC does not buy ASML because the two companies have fundamentally different roles in a specialized, interdependent supply chain rather than competing against one another. [1, 2]

• Different Core Businesses: ASML designs and builds extreme ultraviolet (EUV) lithography machines (the "printing presses" of the chip world), while TSMC is a contract manufacturer (foundry) that operates those machines to produce physical chips for designers like Apple and Nvidia. [1, 2]
• The "Customer Conflict" Rule: As discussed by users on Reddit, major foundries avoid vertical integration into toolmaking because it would mean competing with ASML's other major clients, such as Intel and Samsung, which would likely cause those customers to abandon TSMC. [1, 2]
• Massive Divergence in Expertise and Cost: Running a multi-billion-dollar R&D and precision-engineering firm that builds lithography systems requires an entirely different technical competency and capital structure than operating massive cleanroom fabrication plants. [1]
• A Symbiotic Partnership Exists Instead: Rather than buying each other out, TSMC, Intel, and Samsung previously took minority stakes in ASML through a co-investment program to help fund early EUV development, relying on a collaborative ecosystem rather than ownership. In fact, TSMC has historically been selective and cautious about purchasing ASML's newest, most expensive machinery (such as High-NA EUV) when existing tech makes better economic sense for their nodes. [1, 2, 3]

If you'd like, let me know:

• Are you interested in how chip fabrication economics work?
• Or would you like to explore ASML's export controls and geopolitical role?

AI responses may include mistakes. Learn more

I've always wondered why ASML, TSMC, or Nvidea can't ...

Does TSMC specify to ASML which kind of machines they would prefer

TSMC doesn’t just prefer certain machines; they heavily dictate, customize, and co-develop the machinery ASML builds. [1]

The relationship between the two companies is a tight feedback loop. Because TSMC is ASML’s largest customer, ASML designs its product roadmap directly around what TSMC needs to manufacture next-generation chips. [1, 2, 3]

TSMC steers ASML's engineering in several ways:

• Massive Order Customization: No two ASML lithography machines shipped to a fab are identical. When TSMC plans its upcoming process nodes (like N2 or A16), it maps out its requirements years in advance. They select from over 30 client-specific configuration options to tailor the light sources, throughput, and overlay accuracy of each Deep Ultraviolet (DUV) and Extreme Ultraviolet (EUV) machine. [1, 2]
• The "Too Expensive" Leverage: TSMC holds immense pricing power and uses it to push ASML toward economic efficiency. For instance, TSMC openly pushed back on adopting ASML’s initial High-NA EUV machines (the EXE line) for their upcoming A13 node, calling the $400 million price tag too high for the performance yield. By stalling orders, TSMC forced ASML to figure out how to make the technology more commercially viable. [1, 2, 3]
• Direct Industry Steering (The 12-Inch Mask Initiative): An excellent example of TSMC specifying what it wants is the joint initiative to transition the entire industry from traditional 6-inch photomasks to 12-inch photomasks. High-NA EUV optics shrink the exposure field, which means current tools would have to print massive AI chips in halves and "stitch" them together. TSMC pushed ASML to re-engineer the system for larger masks to avoid this bottleneck, boosting system productivity by a projected 40%. [1, 2, 3]
• Co-Investment and Shared R&D: Historically, TSMC has literally funded the specific technical pathways they wanted ASML to explore. During early EUV research, TSMC bought minority shares in ASML specifically to fund the research tracks that would serve TSMC's proprietary "immersion lithography" and early EUV roadmaps. [1]

Rather than just choosing items off a shelf, TSMC acts as the master architect, telling ASML exactly what dimensions, throughputs, and economic thresholds their future manufacturing tools must hit. [1]

Would you like to dive deeper into the 12-inch photomask transition and how it impacts AI chip sizes, or explore the economic standoff between TSMC and ASML over High-NA machine costs?

AI responses may include mistakes. Learn more

• 

  Manufacturing Digital

  Semiconductors: TSMC is Not Buying ASML's New Machines

  TSMC has decided not to buy ASML's new extreme ultraviolet lithography machine, causing the Dutch company's stock price to sink US...

  
• 

  CNBC

  TSMC, Samsung to use ASML High NA EUV tools for AI chips

  8 Sept 2026 — TSMC expects to adopt the technology in 2030 for advanced chips. * ASML's cost around $400 million,

  
• 

  Forbes

  TSMC To Use ASML Photolithography Gear For Next-Gen Chips

  8 Sept 2026 — Global chip giant TSMC is investing in ASML's cutting-edge EXE lithography machines, essential for manufacturing advanced AI chips...

  

Show all

Of many people would it take at minimum to design a high end integrated circuit

Designing a high-end integrated circuit (IC) from scratch—such as a modern Apple M-series chip, an Nvidia AI GPU, or an Intel/AMD CPU—requires an absolute minimum of roughly 300 to 500 engineering specialists, though a brand-new architecture routinely requires 800 to 1,000+ people. [1, 2]

While a tiny team of 2 or 3 engineers can design a simple, low-end application-specific integrated circuit (ASIC) or an analog chip by relying on heavily pre-made blueprints, a "high-end" chip has billions of transistors and requires a massive, hyper-specialised army. [1, 2]

The minimum team size is determined by the distinct engineering disciplines required across the chip design pipeline.

The Minimum Specialized Sub-Teams Required

If you tried to build a high-end chip with anything less than a few hundred people, the project would stall because one engineer cannot master all the highly fragmented layers of the modern design stack:

• Architecture & Logic Design (30–50 people): These high-level engineers define the chip's behavior and layout rules. They write the code (usually in languages like Verilog or VHDL) that acts as the chip's logical blueprint. [1]
• Design Verification (150–250 people): This is the largest team by far. As users on Reddit discuss, verification engineers scale massively because high-end chips cannot have mistakes before they are manufactured. A single bug can ruin a $50 million manufacturing run (tape-out). Verification typically consumes 40% to 60% of the entire project's human resources. [1, 2, 3]
• Physical Design & Layout (50–100 people): These engineers take the text-based logic code and map it into a physical 2D/3D layout of transistors and wires. They ensure billions of signals do not physically cross, overheat, or experience quantum interference at atomic scales. [1, 2]
• Design for Test (DFT) & Post-Silicon Validation (30–50 people): They build hidden diagnostic pathways into the chip so that when physical silicon comes back from the foundry (like TSMC), they can test it for structural defects. [1]
• Software, Tools & Infrastructure (20–40 people): High-end chips are too complex for humans to map out by hand. Teams require engineers solely dedicated to managing Electronic Design Automation (EDA) software, writing custom AI placement scripts, and running massive supercomputer simulations. [1, 2]

Why You Can't Lean on a Smaller Team

Modern chip design is a brutal optimization game between Power, Performance, and Area (PPA). If a company attempts to use a small, skeleton crew for a high-end chip, two things happen:

1. The "Time-to-Market" Failure: Chip nodes evolve rapidly. If a team of 50 people tried to build a modern GPU, it would take them a decade to verify and layout the billions of logic blocks. By the time the chip was finished, the technology would be completely obsolete.
2. The Third-Party IP Caveat: The only way to marginally shrink a team is by buying pre-designed blocks of code (Intellectual Property/IP) from companies like ARM or Synopsys for standard parts of the chip (like PCIe slots, HDMI controllers, or basic memory interfaces). However, designing the proprietary, high-end core that makes the chip unique still requires hundreds of dedicated specialists. [1, 2]

If you're curious about how this process works, would you like to explore why verification takes up the most manpower, or look into the multi-million dollar software and tool costs required to back these engineering teams?

AI responses may include mistakes. Learn more

• 

  Quora

  If it takes nearly thirty years to count to a billion, one second at ...

  June 27, 2021 — Humans design complex CPUs and graphics cards with billions of transistors by not designing each transistor individually. This is ...
• 

  Medium·Arpita Paul

  Building a chip from Scratch — What it takes to actually build a ...

  Verification is the most time-consuming and labor-intensive process in chip design. According to ARM 40% of a project's resources ...
• 

  ALLPCB

  Why Is Chip Design So Difficult?

  Chips require intensive capital and advanced technical capabilities, passing through design, fabrication, packaging, and testing; ...

  

Show all

Are cutting edge IC's ever designed the open source way

No, truly cutting-edge integrated circuits (ICs) are almost never designed in a fully open-source way.

While the open-source hardware movement has grown rapidly, there is a massive divide between "academic/hobbyist" open source and the "cutting-edge" commercial silicon engineered by companies like Nvidia, Apple, and AMD. [1, 2]

To understand why, it helps to break a chip's creation down into three distinct layers: the Blueprints (IP), the Software Tools (EDA), and the Manufacturing Kits (PDK). [1]

1. The Blueprints (IP): Where Open Source Succeeds

The one area where open source has successfully reached high-end computing is at the blueprint layer, driven entirely by RISC-V. [1]

• What is open: RISC-V is a free, open-standard Instruction Set Architecture (ISA)—essentially an open-source "vocabulary" for processors. [1]
• The Catch: RISC-V is just a specification sheet. While a company like Qualcomm or Ventana Micro Systems can take RISC-V and design an elite server chip, their actual microarchitecture code, logic layout, and safety verification remain entirely proprietary and secret. True "open-source" chip blueprints, like the Linux-capable Basilisk chip, are restricted to legacy manufacturing technology. [1, 2]

2. The Software Tools (EDA): The Multi-Million Dollar Barrier

To translate a chip's logic into physical wiring, engineers must use Electronic Design Automation (EDA) software. [1]

• Excellent open-source EDA initiatives exist (like the OpenROAD Project or OpenLane).
• However, open-source EDA tools max out around 90nm to 130nm technology nodes.
• For a cutting-edge 3nm or 2nm AI chip, open-source tools lack the automation and simulation precision required. Designers are forced to pay tens of millions of dollars for closed-source, proprietary software from giants like Synopsys or Cadence. [1, 2, 3, 4, 5]

3. The Manufacturing Kits (PDK): The Trade Secret Wall

A Process Design Kit (PDK) is the digital rulebook provided by a foundry (like TSMC) that tells design software exactly how to draw transistors without violating physical manufacturing limits. [1, 2]

• The Open-Source Node Cap: Foundries like SkyWater and GlobalFoundries have released fantastic, fully open-source PDKs (like SKY130), allowing anyone to manufacture a chip for free or cheap. [1, 2]
• The Commercial Wall: Advanced foundries treat their cutting-edge PDKs (anything below 14nm down to 2nm) as highly confidential intellectual property and nation-level trade secrets. You cannot get a TSMC 3nm PDK without signing intense non-disclosure agreements (NDAs). This completely illegalises true open-source design at the cutting edge, because you cannot publicly share a design that contains a foundry's secret manufacturing math. [1]

Summary: The Two Universes of Chip Design

The landscape remains strictly divided between two models:

Attribute	Fully Open-Source Way	Cutting-Edge Commercial Way
Foundry Nodes	Legacy (130nm to 180nm)	Frontier (5nm to 2nm)
Software Costs	Free (OpenROAD, KLayout)	$10M+ Enterprise Licenses
Core Architecture	Fully open, shared code	Proprietary, fiercely guarded
Typical Target	Microcontrollers, IoT, Academia	AI Supercomputers, Mobile SoC, Workstations

If you'd like, let me know:

Do you want to explore how RISC-V development boards like the VisionFive 2 let hobbyists experiment with open architecture?Or would you like to know more about the national security reasons behind why advanced PDKs are kept top secret?

AI responses may include mistakes. Learn more

• 

  

  YouTube·RISC-V International

  Open Source Chip Design in the European Semiconductor ...

  28 May 2025 — open source process design kits (PDKs) Open source PDKs are crucial for manufacturing, manufactured using a 130 nanometer process ...

  

  1m
• 

  Princeton University

  Open-source platform breaks a bottleneck for high

  David Wentzlaff have developed an open source platform to make high-tech chips easier and cheaper to design, especially for acadmi...

  
• 

  University of Michigan

  Open-source hardware: a growing movement to democratize IC design

  providing open-source process design kits (PDKs), EDA tools, and free fabrication via Google, NIST, SkyWater Technology (130 nm),

  

Show all

Is the what is the 'Basilik' chip good for, what can it do

The Basilisk chip is a landmark research microchip designed by researchers at ETH Zürich and the University of Bologna. [1, 2]

What makes it unique is that it is the world's first end-to-end, fully open-source microchip capable of autonomously booting the Linux operating system. Every single line of its hardware blueprint, the software tools used to design it, its boot code, and its underlying manufacturing kit are completely open-source and auditable. [1, 2]

It is vital to understand that Basilisk is not a commercial product designed to compete with your phone or PC. It is a "research demonstrator" engineered to prove that the world can build highly secure, auditable, and sovereign computer platforms without relying on a single piece of secret corporate intellectual property (IP). [1, 2]

What Can the Basilisk Chip Do?

Despite being built on an older, inexpensive 130nm manufacturing node, Basilisk has enough hardware integrated directly onto its 34 mm² silicon die to run as a fully functional, self-contained miniature computer. It can: [1, 2, 3]

• Boot a General-Purpose Operating System: It runs a full distribution of 64-bit Linux from the moment it boots up. [1, 2]
• Drive Video and Graphics: It includes a custom built-in display engine and physical VGA output, meaning it can output a visual desktop or terminal interface to a standard computer monitor. [1, 2]
• Interact with Standard USB Devices: It features integrated USB 1.1 host controllers, allowing you to plug in and use basic computer peripherals like keyboards and mice. [1, 2]
• Manage External Dynamic Memory: It embeds a digital HyperRAM controller to manage up to 128 MB of external memory, which provides the necessary workspace to load the Linux kernel and handle basic computational workflows. [1, 2]
• Communicate via High-Speed Connections: It includes a custom 62 Mb/s chip-to-chip (C2C) interface that lets it send and receive data smoothly from neighboring processors or hardware boards. [1, 2]

What Is It Good For? (Its Real-World Purpose)

Because Basilisk runs at a modest peak clock speed of 62 MHz to 102 MHz, it isn't meant for gaming or intense web browsing. Instead, it serves critical roles in high-security and academic fields: [1]

• "Zero-Trust" Government and Military Security: Standard chips contain proprietary "black box" components that could theoretically harbor hidden hardware backdoors or vulnerabilities. Because Basilisk's blueprints are 100% transparent, security agencies can audit every transistor to guarantee there are no hidden tracking or spying mechanisms.
• Supply Chain Sovereignty: Countries or organizations can use the Basilisk design to manufacture their own essential computing infrastructure completely independently of mainstream tech monopolies or geopolitical export bans.
• Democratizing Semiconductor Education: Before Basilisk, students and researchers had to rely on "academic toys" or sign multi-million dollar corporate non-disclosure agreements (NDAs) to study real application-class silicon. Basilisk provides a freely accessible, industrial-grade blueprint of a working computer chip that universities worldwide can replicate and study. [1, 2, 3, 4]

If you would like to know more, let me know:

• Would you like to explore how the open-source tools (like Yosys and OpenROAD) physically mapped out this chip?
• Or are you curious about RISC-V architecture and why it's dominating the open-source hardware movement?

The debate over whether it is "unfair" for Big Tech to profit from open-source software is one of the most contentious and complex issues in modern technology. There is no universal consensus; instead, the industry is deeply divided, and the conversation involves sharp trade-offs between ethical fairness, commercial reality, and the literal text of open-source licenses.

The issue is best understood by looking at the two primary, conflicting arguments.

---

The Argument That It Is Unfair (The "Parasitic" View)

Many maintainers, independent developers, and smaller software companies argue that Big Tech's exploitation of open source is highly extractive and unsustainable.

• Value Extraction Without Contribution: Critics point out that hyper-scalers (like AWS, Google, and Microsoft) pull freely available code, package it as a managed cloud service, and generate billions of dollars in revenue without necessarily contributing code, bug fixes, or financial resources back to the original project.
• The "Strip-Mining" Effect: When a cloud giant offers a managed version of an open-source tool, it often dries up the revenue of the commercial company that actually employs the tool's core maintainers (e.g., how AWS historically competed with Elastic or MongoDB). This has driven many open-source companies to abandon true open-source licenses for restrictive "source-available" licenses.
• Developer Burnout: Individual maintainers often work for free on projects that secure global infrastructure. When Big Tech uses this code at scale, developers face immense pressure to patch critical vulnerabilities (like the Log4j crisis) under stressful conditions, receiving little to no compensation from the corporations benefiting from their free labor.

---

The Argument That It Is Fair (The "System Works" View)

Conversely, many legal scholars, enterprise advocates, and even prominent open-source purists argue that Big Tech is operating exactly as open-source software was designed to allow.

• Compliance with the License: The fundamental definition of open source (under licenses like MIT, Apache 2.0, or BSD) explicitly permits commercial use, modification, and distribution for profit. Big Tech is not stealing; they are legally playing by the exact rules the creators agreed to.
• Infrastructure and Distribution Scale: Tech giants provide massive engineering resources to host, scale, secure, and popularize these tools. By integrating open-source projects into their cloud ecosystems, they expose them to millions of new users, effectively making the software an industry standard.
• Massive Reciprocal Contributions: While critics highlight instances of exploitation, tech giants are also some of the largest contributors to the open-source ecosystem. Projects like Linux, Kubernetes (Google), React (Meta), and VS Code (Microsoft) were either created by or are heavily sustained by Big Tech engineers who are paid corporate salaries to write open-source code.

---

Direct Comparison: The Open-Source Paradox

Perspective	Core Philosophy	Main Benefit	Main Risk
The Maintainer's View	Open source is a collaborative community built on mutual respect and shared burden.	Democratises technology and prevents corporate monopolies.	Financial ruin and burnout for creators while corporations profit.
The Corporate View	Open source is a legal framework designed to maximize software adoption and utility.	Provides unparalleled infrastructure scale, security audits, and developer adoption.	Fragmenting the community if licensing conflicts force codebases to split.

---

The Industry's Response: Finding a Middle Ground

Because this friction has reached a boiling point, the tech ecosystem has begun actively shifting to protect creators through several mechanisms:

• License Relicensing: Companies like MongoDB, Redis, and Elastic famously moved away from standard open-source licenses to "source-available" licenses (like the SSPL or BSL). These licenses state that anyone can use the code for free except for cloud providers trying to sell it as a managed service.
• The Rise of "Fair-Code" Initiatives: New funding models, such as GitHub Sponsors, Open Collective, and corporate grants, are attempting to funnel money directly from enterprise budgets back to the specific developers maintaining their dependencies.

Ultimately, whether it is "unfair" depends on your vantage point. Legally and economically, it is a feature of the system. Ethically and sustainably, it is a flaw that the tech industry is actively trying to patch.

Are you asking this from the perspective of an independent developer trying to monetize a project, or are you looking at how licensing changes (like those from Redis or Elastic) impact enterprise cloud architectures?

AI responses may include mistakes. Learn more

I'm looking at it from an anti-trust perspective and wondering whether money could be spread around more, I guess

Looking at this from an antitrust and wealth redistribution perspective shifts the conversation from a simple "licensing dispute" to a structural question: How can the massive financial gains concentrated at the top of the tech ecosystem be reallocated to sustain the foundational infrastructure underneath it?

Applying traditional antitrust principles to open source reveals unique challenges, but it also opens the door to potential mechanisms for spreading the money around more equitably.

---

1. The Antitrust Dilemma: A New Form of "Monopsony"

Historically, antitrust law (like the Sherman Act) focused on monopolies—where a single giant overcharges consumers. But with open source, Big Tech exhibits what economists call monopsony or "extractive" market power: [1, 2]

• The Infrastructure Trap: Companies like AWS, Microsoft, and Google don't overcharge for open-source software; they provide it to the public for free or cheap. However, they act as the dominant gatekeepers and distribution channels. [1, 2]
• Capturing the Value: Because they control the cloud infrastructure, the profits flow entirely to their compute, hosting, and managed services. The independent creators who built the actual software logic are starved of revenue because they lack the capital to compete with Big Tech's distribution networks. [1]

---

2. Can Money Be Spread Around More? (Proposed Mechanisms)

If regulators or the industry were to tackle this structurally, there are several ways the wealth could be reallocated:

A. The "Digital Tax" or Revenue-Share Mandate

Some economic and legal theorists propose a dual-licensing framework with a revenue threshold. [1]

• How it works: Software remains completely free for individual developers, startups, nonprofits, and academics. However, if a corporation's revenue exceeds a certain threshold (e.g., $1 billion) and they commercially exploit that software as part of their core infrastructure, they would be legally required to pay a micro-percentage (e.g., 0.15%) of that specific revenue back to a neutral, independent fund. [1]
• The Antitrust Angle: This acts as an "interconnection fee," similar to how telecom giants historically had to pay smaller networks to route calls. It forces the dominant platform to pay for the upstream resources it relies on to maintain its monopoly.

B. Direct Developer Funding (Bypassing "Captured" Foundations)

Currently, Big Tech gives millions of dollars to open-source consortia, such as the Linux Foundation. However, critics argue these foundations have been "captured" by their largest corporate donors. [1]

• The Misallocation: A massive portion of that corporate funding goes toward administrative overhead, marketing, tech conferences, and executive salaries, rather than directly into the pockets of the independent developers writing the code.
• The Solution: A true redistribution framework would bypass these corporate-aligned middle-men. Platforms like Open Collective or automated protocols could track code dependencies and mandate that corporate tax-deductible contributions go directly into the digital wallets of core maintainers based on how heavily their libraries are imported. [1]

C. Structural Separation (Breaking Up the Cloud)

If the U.S. Department of Justice (DOJ) or Federal Trade Commission (FTC) pursued a structural breakup of Big Tech—such as separating Amazon's retail platform from AWS, or Google Search from Google Cloud—it would indirectly help open source. [1, 2]

• Levelling the Playing Field: By severing the cloud hosting platform from the mega-corporation’s infinite capital, smaller, specialized open-source companies (like Elastic or MongoDB) could compete on a level playing field to sell their own managed cloud services without facing an opponent with deep, monopolistic cross-subsidies.

D. The "Fair Source" Movement

A compromise rising within the developer community is the Fair Source framework. Unlike strict open source (which allows anyone to do anything), Fair Source licenses explicitly state: "You can read, modify, and use this code for free, but if your company scales past a certain number of employees or users, you must buy a commercial license." [1]

• While purists argue this destroys the original ethos of free software, proponents see it as a necessary form of digital wealth redistribution to keep independent innovation alive. [1]

---

The Reality Check

Right now, antitrust law is struggling to keep pace. Courts still rely heavily on the "consumer welfare standard," which says that if a corporate practice keeps prices low or free for the end consumer, it isn't an antitrust violation. Because open source keeps software cheap for the public, traditional courts rarely see Big Tech’s extraction of value as a "harm," ignoring the systemic burnout and collapse occurring behind the scenes. [1, 2]

Changes will likely have to come from legislative updates to antitrust laws or a unified, collective shift in how developers license their code to legally force the hands of hyper-scalers. [1, 2]

Do you think a government-mandated "tax" on enterprise cloud usage is the right approach to fund these developers, or do you think the solution should come from new software licenses that developers choose themselves?

AI responses may include mistakes. Learn more

How Big Tech Captured Open Source Governance

Can you explain the 'source-available' model

The source-available model is a software licensing approach where a creator makes their source code publicly visible, readable, and modifiable, but places strict legal limits on how it can be commercially used or monetised.

It bridges the gap between traditional proprietary software (where code is hidden) and true open-source software (where code is completely free to use for any purpose).

---

Why It Exists: The "Big Tech" Shield

The model exploded in popularity because open-source companies realized that cloud hyper-scalers (like Amazon Web Services) were taking their free code, packaging it as a managed cloud service, and making billions of dollars without contributing back.

By switching to a source-available license, software creators can say: "You can use our software for free in your own apps, but you cannot copy our code to build a competing cloud service."

---

Source-Available vs. Open Source vs. Proprietary

To understand how it functions, it helps to compare it directly against the other major software models:

Feature	Open Source (e.g., Apache 2.0, MIT)	Source-Available (e.g., BSL, SSPL)	Closed-Source / Proprietary
Can you view the code?	Yes	Yes	No
Can you modify the code?	Yes	Yes	No
Is it free for normal business use?	Yes	Yes (Usually)	No
Can you use it to sell a competing cloud service?	Yes	No	No
Who approves the license?	Open Source Initiative (OSI)	The Creator / Company	The Corporation

---

How the Model Works in Practice

Most source-available licenses follow a set of predictable mechanics to balance developer goodwill with corporate protection:

• The Cloud Exclusion Clause: You can run the software for your own company, host your own database, and build your own applications entirely for free. The only thing you cannot do is sell that software "as-a-service" to others.
• The Delayed Open-Source Clause: Many licenses use a "convertible" mechanic (like MariaDB’s Business Source License). The code is source-available for a set period (e.g., 3 to 4 years), after which that specific version automatically converts into a fully free, Open Source license (like Apache 2.0).
• The Scale Trigger: Some licenses are free up to a certain point—such as a specific revenue threshold or number of employees—after which an enterprise license must be purchased.

---

Real-World Examples of the Shift

Several major tech infrastructure companies have famously abandoned true open source in favor of the source-available model to protect their businesses:

• Redis: In 2024, the popular in-memory database shifted away from the open-source BSD license to a dual source-available model (RSALv2 and SSPLv1) to prevent cloud providers from offering hosted Redis without commercial agreements.
• Elastic (Elasticsearch): Moved to a source-available model in 2021 after a fierce market battle with AWS over hosted search engines (though they later re-introduced open-source options once market protections were established).
• MongoDB: Created the Server Side Public License (SSPL), which mandates that if you offer MongoDB as a publicly available service, you must release the entire source code of your hosting infrastructure—a requirement cloud giants refuse to meet.

---

The Core Trade-Off

The source-available model is highly polarizing in the tech community.

• The Pros: It keeps the code visible for security audits, allows independent developers and startups to build on powerful tools for free, and creates a sustainable business model so software creators don't burn out or go bankrupt.
• The Cons: It fragments the developer ecosystem. True open-source purists argue that source-available software should never be called "open" because it restricts freedom, and it forces larger enterprises to navigate complex legal compliance checks to ensure they aren't accidentally violating usage clauses.

Engineers do literally "code" physical microchips using specialized languages. Instead of traditional software, they write text using Hardware Description Languages (HDLs) like Verilog or VHDL to outline how a chip's logic gates must behave. [1]

How Chip Coding Works

Behavioral Code: Designers write plain-text HDL code that defines the electrical and logical functions of a processor or accelerator.
Logic Synthesis: Software tools translate that code into a vast, interconnected network of physical logic gates.
Layout & Tape-out: Automated tools pack those gates into a dense physical floor plan, routing every microscopic wire before sending the design to a silicon fabrication plant. [1, 2]

The Shift to AI-Assisted Design

AI-Generated Chips: Systems like OpenAI and Synopsys's GPT-Synopsys or NVIDIA Nemotron have started automating and co-coding RTL (Register-Transfer Level) hardware blocks, cutting design cycles dramatically (such as OpenAI's Jalapeño chip reaching tape-out in 9 months). [1, 2, 3]
Bespoke Architectures: Researchers at Princeton have even used reinforcement learning to have AI design non-intuitive, high-frequency RF chips that outperform human layouts. [1, 2]
Simplicity First: As top tech leaders like Jensen Huang note, the best engineering often focuses on clean, minimal logic rather than over-complicated code—a philosophy that applies whether you are writing software or synthesizing silicon. [1, 2]

You can read more about software engineering philosophy on Medium. [1]