All my new code will be closed-source from now on

Original Article ↗ Hacker News Discussion ↗

Political / economic framing

  • Some argue that both “Ludditism” and attempts to abolish “the system” have failed: industrial capitalism and automation delivered higher productivity and removed many undesirable manual jobs.
  • Others counter that this caricatures Marxism: in that view, automation is a precondition for socialism, not something Marxists oppose.
  • Thread debates whether current state-capitalist regimes (e.g. China) are genuinely Marxist or simply exploit socialist rhetoric while using capitalist-style enclosure and rent-seeking.
  • Several comments stress that core tech (computers, internet, web) and open source arose largely from public research and non-market motives; capital then fenced and monetized them.

Why people do OSS

  • Many participants say they open source code primarily to help others or for ideological/educational reasons, not for direct income.
  • Others report indirect benefits: easier hiring, reputation, consulting.
  • Strong disagreement with any framing that treats all OSS participation as fundamentally money-motivated.

Monetization models and their limits

  • Examples discussed: donations, employer-sponsored time, support/consulting, “open core” with paid premium features, SaaS around an OSS core.
  • Some note large projects (kernel, major DBs, Wine) are funded by companies paying for needed features.
  • Others claim that, in practice, companies mostly free-ride: “pip install” everything, forbid contributions, and treat OSS as a cheap supply chain.

AI/LLMs as a breaking point

  • Many see LLMs as intensifying extraction: models are trained on OSS and docs, then used to generate code and answers without attribution or traffic, eroding maintainers’ ability to monetize.
  • Concern that “LLM-optimized docs” make agents better while reducing human visits and revenue.
  • Some argue AI can already recreate “premium” features from open-core projects, undermining that model; others think complex premium features remain hard to auto-generate.

Licensing, free software vs open source

  • Distinction drawn between “free software” (freedom/ideology) and “open source” (often pragmatic, company-friendly).
  • Worry that AI effectively bypasses copyleft (e.g. GPL/AGPL) by learning from code while ignoring license obligations.
  • Some insist OSS is inherently charitable; expecting guaranteed income from it is a category error.

Tailwind and concrete cases

  • Tailwind is cited as heavily used yet seeing reduced revenue and layoffs; some blame AI, others say donations are up and the real issue is weak proprietary upsells.
  • One view: Tailwind solved “CSS complexity,” and AI now solves the same problem; another replies that LLMs increase Tailwind usage rather than replace it (data details are disputed/unclear).

Proposed responses and new mechanisms

  • Suggestions: source-available with paid commercial licenses; metered API-style libraries “for agents”; stronger OSS licenses limiting AI/commercial use; standardized donation/funding flows via tooling (e.g. “fund all deps with one command”); escrow-based feature requests.
  • Others propose public/academic funding or quasi-public corporations to sustain critical software.

On going closed-source

  • Some sympathize with maintainers going closed-source after feeling exploited by hyperscalers and AI vendors.
  • Others predict that paywalled libraries may simply lose adoption, especially if AI can reimplement their value.
  • Several insist their own commitment to FOSS is unchanged: they need transparency, modifiability, and community collaboration regardless of AI.

Side thread: replacing git / forges

  • One participant describes an OSS effort to “replace git/GitHub” with semantic, AST-based version control and web-native contribution workflows, arguing openness is a competitive weapon.
  • Skeptics see git as “peak human-centered source control” and question the viability of displacing it, especially without LLM-focused advantages.