AI makes software cheaper to produce. It does not make good software free to maintain.
That distinction matters. Generating a prototype may take minutes, but the durable asset still needs a clear purpose, trustworthy dependencies, tests, documentation, security response, distribution, user support, and someone willing to remain accountable when the exciting part ends.
The epoch of change is therefore not “AI replaces software work.” It is a shift in where the scarce work lives.
What AI makes abundant
AI coding systems reduce the cost of:
- translating an idea into a first implementation;
- exploring unfamiliar libraries;
- producing routine tests and documentation;
- porting repetitive code;
- comparing design options;
- finding likely defects.
This is real leverage. A small team can attempt projects that once required a much larger budget.
Abundance at the implementation layer, however, increases the amount of code competing for attention. Judgment, verification, and stewardship become more valuable precisely because generation is cheap.
What remains scarce
A durable software project still requires:
- problem selection: choosing a need worth solving;
- taste: deciding what not to add;
- verification: proving that the system works outside a demo;
- trust: handling security reports and releases responsibly;
- distribution: reaching users who benefit;
- maintenance: adapting when dependencies and platforms change;
- ownership: keeping incentives aligned over time.
AI can assist each task. It cannot make the social obligation disappear.
The open-source funding problem
Free software gives users extraordinary freedom, but maintainers still need food, time, equipment, and protection from burnout. Advertising can distort a product toward surveillance and engagement. Subscriptions can turn every tool into permanent rent. Venture capital can force growth where a small stable utility would have been enough.
There is no universal replacement. A healthy project may combine:
- paid hosting for users who value convenience;
- one-time licenses for polished distributions;
- support and integration contracts;
- grants and public-interest funding;
- sponsorships;
- value-for-value payments;
- cooperative ownership;
- paid complementary services around an open core.
The key question is not whether money enters the project. It is whether the revenue model strengthens or corrupts the user relationship.
Where Bitcoin fits
Bitcoin does not solve maintenance automatically. It does provide useful properties for a global funding layer:
- small or large payments without requiring the creator and supporter to share a bank;
- direct settlement to a self-custodied wallet;
- programmable payment flows through Lightning;
- a scarce asset in which long-term project reserves can be held;
- a culture of value-for-value exchange.
It also brings volatility, custody risk, tax obligations, and usability friction. A project that accepts Bitcoin still needs budgets, accounting, governance, and a reason for people to pay.
The strongest role for Bitcoin here is not speculative salvation. It is credible ownership and direct patronage.
From labor to assets
AI changes the economic goal for creators. If implementation becomes abundant, selling hours becomes less defensible. The better strategy is to convert effort into durable assets:
- reusable software;
- a trusted brand;
- an audience relationship;
- data or research with provenance;
- operational knowledge;
- recurring service revenue;
- reserves that extend the maintenance runway.
An asset keeps producing value after the initial work. That is the bridge between open collaboration and economic survival.
A practical funding design
For a new open project:
- Define the user and recurring problem.
- Release the smallest version that proves utility.
- Make self-hosting and data export real, not ceremonial.
- Identify the ongoing costs honestly.
- Charge for convenience, assurance, or service—not artificial captivity.
- Keep a public boundary between community governance and paid obligations.
- Automate maintenance, but require evidence before release.
- Hold enough runway to survive a bad quarter.
- Measure whether revenue improves the product users came for.
AI lowers the cost of reaching step two. It does not remove steps three through nine.
The future worth building
The most hopeful outcome is not a world of infinite disposable applications. It is a world in which small groups can build excellent tools, users can leave without losing their data, maintainers can earn without turning attention into inventory, and capital can flow directly to work people value.
Open source supplies the freedom. AI supplies leverage. Bitcoin can supply an ownership and payment rail. Human stewardship is what turns the three into durable institutions.
For the broader monetary argument behind this view, read Why Bitcoin changed how I think about money.