1. The strong objection is real
The strongest critique is not “technology is bad.” It is that technology often arrives through unequal institutions. The tools may be digital, but the power relations are old: capital, infrastructure, language, patents, cloud compute, finance, platforms, and states.
The World Bank’s World Development Report 2016: Digital Dividends made exactly this point. Digital technologies can promote growth, jobs, and services, but the benefits are uneven unless societies also build “analog complements”: competition, skills, and accountable institutions.1 That is a restrained institutional version of the same concern: digital tools do not automatically liberate anyone.
The digital divide remains material. ITU’s 2025 data estimate that about 6 billion people are online while roughly 2.2 billion remain offline; low-income countries lag badly in quality, affordability, skills, and 5G coverage.2 AI readiness is similarly uneven. The IMF’s AI Preparedness Index evaluates countries across digital infrastructure, human capital and labor policies, innovation and integration, and regulation and ethics; IMF leadership has warned that developing countries lag badly on the foundations needed to benefit from AI.3
The critique is therefore correct up to this point: without a change in architecture, governance, and ownership, the Global North likely captures disproportionate gains from AI, platforms, data, cloud infrastructure, and digital finance.
2. Climate change changes the denominator
Climate change is not a side issue. It is the background condition under which all digital policy will operate for the next twenty years.
The IPCC says roughly 3.3 to 3.6 billion people live in contexts highly vulnerable to climate change, and that vulnerability is shaped by poverty, governance, conflict, colonialism, and inequity.4 The IPCC also reports that mortality from floods, droughts, and storms between 2010 and 2020 was about 15 times higher in highly vulnerable regions than in regions with very low vulnerability.4 The World Bank’s Groundswell work projects that climate change could force up to 216 million people to move within their own countries by 2050 without urgent climate and development action.5
Climate therefore increases the probability of bad outcomes: food stress, water stress, migration pressure, conflict, authoritarian emergency powers, scapegoating, surveillance borders, and debt. It also increases the value of open digital tools: early-warning systems, resilient communications, climate-risk maps, mobile cash transfers, remittances, telemedicine, independent reporting, local-language education, and coordination during disasters.
| Climate effect | Digital danger | Digital freedom use case |
|---|---|---|
| Disaster | Shutdowns, rumor cascades, platform dependency. | Open communications, independent crisis maps, resilient mesh and satellite links, encrypted coordination. |
| Migration | Biometric tracking, border AI, financial exclusion. | Portable identity with selective disclosure, self-custody, remittances, rights documentation. |
| Food/water stress | Data extraction by agritech monopolies. | Open climate data, local-language advisories, farmer-owned data cooperatives. |
| Authoritarian emergency powers | Censorship, surveillance, frozen accounts. | Encryption, anonymity, censorship-resistant publishing, peer-to-peer payments. |
| AI/compute growth | Energy and water demand concentrated in wealthy regions. | Transparent environmental accounting, clean-energy requirements, local compute commons, efficient open models. |
The environmental footprint of digital technology cannot be ignored. The IEA’s Energy and AI report emphasizes that data centers consumed around 415 TWh of electricity in 2024 and could more than double by 2030, with AI a major driver.6 Bitcoin mining has its own climate and local externalities; Jones, Goodkind, and Berrens estimated substantial climate damages from proof-of-work mining and compared Bitcoin’s climate-damage profile more to “digital crude” than “digital gold.”7
3. The Charlie Brown problem: who is Lucy?
The Charlie Brown football metaphor is useful, but only if the roles are assigned correctly.
The mistake is to say: “Technology promised liberation before, and Lucy pulled the ball away, therefore we should distrust all technology.” The sharper lesson is: do not let Lucy own the field, the ball, the rulebook, and the referee.
Lucy is not “the Internet”
Lucy is platform monopoly, surveillance advertising, state censorship, biometric identity chokepoints, proprietary AI dependency, scam finance, and foreign-owned infrastructure that cannot be audited or exited.
The cypherpunk answer
Move power from institutions to users: open protocols, encryption, self-custody, local control, data minimization, auditable code, portable identity, and speech that does not require permission.
4. A conditional 20-year forecast
No one can forecast this with actuarial precision. The honest answer is conditional. The outcome depends on architecture and governance.
| Path | Likely 2046 outcome | Rough likelihood if current trends dominate |
|---|---|---|
| Captured path | AI, platforms, finance, identity, and cloud infrastructure concentrate in the Global North and local elites. The Global South becomes a data source, labor pool, consumer market, and surveillance frontier. | High risk: roughly 50–60% without serious policy correction. |
| Mixed path | Most people gain useful tools, but benefits are uneven. Education, payments, health, and organizing improve, while monopoly, scams, digital addiction, and surveillance extract much of the upside. | Most likely single path: roughly 35–45%. |
| Self-determination path | Open Internet access, local-language AI, clean compute, privacy, encryption, open-source tools, self-custody, cheap remittances, and accountable platforms increase agency faster than dependency. | Lower unless fought for: roughly 20–30% under current politics; higher with deliberate architecture. |
| Severe bad path | Climate disruption plus digital authoritarianism: biometric control, border AI, censorship, account freezes, monopoly platforms, predatory AI, and low-trust financial systems. | Non-trivial: roughly 15–25% if climate and governance deteriorate together. |
The important point is that these are not fixed laws of nature. They are policy-sensitive. The good outcome is not guaranteed by technology. But the bad outcome is made more likely by closed systems, surveillance identity, censorship, monopoly, and financial gatekeeping.
5. Why speech and digital freedoms matter more under climate stress
In a stable world, free speech is a democratic principle. In a crisis world, it is also an early-warning system.
Climate disruption will make governments more tempted to control information: food shortages, failed flood response, corruption in aid delivery, public-health crises, police abuse, climate migration, and conflict all create incentives to suppress embarrassing facts. Freedom House reports that global Internet freedom has declined for many consecutive years and that people in scores of countries face arrest or imprisonment for online expression.8 Access Now’s #KeepItOn coalition has documented record levels of Internet shutdowns, often during conflict, protest, elections, and crisis.9
Peer-reviewed censorship research shows why this matters. King, Pan, and Roberts found that Chinese censorship often targets collective-action potential more than mere criticism; harsh criticism may be tolerated, while organizing capacity is suppressed.10 Penney’s work on surveillance and chilling effects shows that surveillance and legal threats can suppress lawful speech, search, and sharing.11
In climate-stressed societies, people will need to document water theft, corruption, failed infrastructure, illegal evictions, police violence, border abuses, and unsafe labor conditions. They will need to organize mutual aid and expose official lies. Those activities require speech, privacy, encryption, and anonymity.
6. Cypherpunk principles are social goods
Cypherpunk does not mean “let tech companies do whatever they want.” It means using technical architecture to reduce dependence on institutions that can be corrupt, captured, incompetent, discriminatory, or coercive.
Eric Hughes’s Cypherpunk’s Manifesto defines privacy as selective revelation, not secrecy for its own sake.12 The UN Special Rapporteur David Kaye made a parallel human-rights argument: encryption and anonymity enable freedom of opinion and expression in the digital age, especially for vulnerable speakers.13 Abelson, Gilmore, and coauthors warned in the peer-reviewed Journal of Cybersecurity that mandated exceptional access to communications systems creates security and human-rights risks.14
Encryption
Protects journalists, migrants, women, patients, dissidents, children, lawyers, workers, and families. Weakening it for “safety” weakens it for everyone.
Anonymity and pseudonymity
Let vulnerable people speak, learn, organize, and report abuse without exposing themselves to employers, mobs, abusers, or states.
Open protocols
Reduce dependency on a single platform owner. They allow local builders to adapt tools instead of renting permission from foreign monopolies.
Self-custody
Provides a financial exit when banks, states, payment processors, or families can freeze, censor, or confiscate money.
Data minimization
Reduces breach risk, surveillance risk, and predatory profiling. The safest database is often the one never built.
Verifiability
Open code, public audits, cryptographic proofs, reproducible models, and transparent reserves reduce the need to trust elites.
Nick Couldry and Ulises Mejias’s work on data colonialism warns that digital capitalism can turn human life into raw material for extraction.15 A cypherpunk lens answers that danger by minimizing coerced data capture, preserving exit rights, and refusing identity systems that make every person legible to power by default.
7. AI: tutor or dependency stack?
The concern that AI will replace people after social media has weakened learning is not absurd. Web2 has trained many users into passive consumption. AI can intensify that if it becomes an answer machine, persuasion engine, surveillance layer, or labor-displacement system owned by a few firms.
But the conclusion should be design-specific. Education technology can improve learning when it complements teachers and adapts to students. Muralidharan, Singh, and Ganimian’s randomized evidence on technology-aided instruction in India found meaningful learning gains from personalized, computer-assisted after-school instruction.16 The lesson is not that software replaces education; it is that well-designed tools can augment it.
AI’s risks are real. Bender and coauthors warned that large language models can reproduce bias, generate unreliable fluent text, and concentrate power.17 IMF and UNDP warnings about AI inequality are also real: if compute, data, chips, cloud, and language coverage are concentrated in wealthy countries, AI can widen global gaps rather than close them.318
The cypherpunk-informed position is not “AI at any cost.” It is:
- Local-language models and datasets.
- Open and auditable models for public-interest uses.
- Strict privacy and data-minimization rules.
- Clean-energy and water-accounting requirements for data centers.
- Public compute for schools, researchers, and civil society.
- Human-in-the-loop safeguards in health, law, education, and welfare.
- Anti-monopoly rules so AI is not rented from a few foreign cloud empires forever.
9. Crypto: regulate the casino, preserve the exit
Crypto has deserved much of its criticism: scams, hacks, leverage, FTX-style custodial fraud, memecoin predation, ransomware, and proof-of-work externalities. But dismissing every permissionless financial tool because the industry contains casinos is a one-column analysis.
The real question is comparative. Many people face high remittance costs, inflation, capital controls, weak banking access, sanctions spillovers, de-risking, or account freezes. World Bank Findex data show major progress in account ownership but continuing exclusion.20 The World Bank’s remittance data show that global remittance costs remain far above the Sustainable Development Goal target.21 Peer-reviewed mobile-money evidence from Kenya shows that better financial rails can reduce poverty and improve women’s occupational choices.22
Human-rights Bitcoin advocates such as Alex Gladstein should be understood as moral witnesses, not econometric authorities. Their core point is that money can be a tool of repression: frozen accounts, capital controls, inflation, surveillance, and blocked donations can punish speech and civil society without formally banning either.23
The policy answer is strict where the risks are highest and protective where the civil-liberties value is highest:
Regulate harder
Custodial exchanges, leverage, stablecoin reserves, celebrity promotions, mining externalities, market manipulation, fraud, custody, and cybersecurity.
Protect strongly
Self-custody, open-source wallets, lawful privacy tools, small-dollar remittances, humanitarian transfers, and peer-to-peer financial exit.
10. Policy architecture for the good outcome
The good outcome is not guaranteed. It has to be built. The policy agenda is not “less regulation”; it is less domination.
Ten principles
- Universal affordable Internet access. Treat connectivity as infrastructure for education, health, markets, speech, and disaster response.
- Protect encryption and anonymity. Do not solve safety problems by making every vulnerable person legible to every powerful institution.
- Regulate Web2 design. Audit recommender systems, ban dark patterns, restrict behavioral ads to children, and require transparency and appeals.
- Build local-language AI capacity. Avoid a world where the Global South only rents English-centric models from foreign cloud providers.
- Make AI environmentally accountable. Require data-center disclosure of electricity, water, land, and emissions; align compute growth with clean energy and grid resilience.
- Support open-source public-interest AI. Public compute, open datasets with consent and governance, and local research institutions can prevent total dependency.
- Preserve self-custody and open financial rails. Regulate custodians and fraud while protecting lawful peer-to-peer transactions and remittances.
- Lower remittance costs. Competition, stablecoin reserve rules, faster payments, and consumer protection should serve migrants and families.
- Guarantee data sovereignty and minimization. Communities should not become raw material for extractive AI and advertising systems.
- Defend speech during crisis. No shutdowns, no blanket censorship, no political de-banking, and no emergency powers that silence disaster victims or dissidents.
Conclusion
The concern that the Global North will capture most of the upside is not only plausible; it is the default trajectory if current institutions remain unchanged. Climate change makes that trajectory more dangerous. AI and data centers can deepen energy and water stress. Social platforms can damage politics and attention. Crypto can prey on desperation. Migration and scarcity can tempt states toward surveillance and censorship.
But that is precisely why speech and digital freedom matter. In the world being described — hotter, more unequal, more migratory, more politically volatile — vulnerable people will need ways to communicate, organize, preserve savings, send money, document abuse, expose corruption, learn, and exit broken systems.
The final formulation:
The Charlie Brown answer:
8. Social media: keep organizing power, regulate Web2
Social media is the most mixed case. It can help people organize and document abuse, but Web2 platforms are also optimized for addiction, outrage, manipulation, harassment, and surveillance advertising.
Tufekci and Wilson’s peer-reviewed study of Tahrir Square found that social media helped protesters access information the regime could not easily control and shaped protest participation and logistics.19 But the same networked tools can also be captured by states, bots, extremists, and algorithmic engagement systems.
The right policy is not to censor the public square or impose universal identity checks. It is to regulate Web2’s harmful design: recommender systems, targeted ads, dark patterns, bot networks, doxxing, harassment, algorithmic amplification, and data extraction. Preserve the ability to organize. Regulate the attention casino.