Impact of AI on cryptoanarchy

How artificial intelligence reshapes the battle between individuals and states

Pavol Lupták, iam@opportunist.global

Software

The threat:

AI is the most powerful surveillance and control technology ever invented.

It will be deployed against you by every state that can get its hands on it.

Software

Timothy C May, 1988:

"A specter is haunting the modern world, the specter of crypto anarchy."

"The State will, of course, try to slow or halt the spread of this technology, citing national security concerns, use by drug dealers and tax evaders, and fears of societal disintegration."

Replace "this technology" with "artificial intelligence" — notice anything?

Software

The four Horsemen of the AI State

(reference to Timothy May's Four Horsemen Of The Infocalypse )

1. CBDC (programmable money, e.g., digital euro with behavioral triggers controlled by AI)

2. Chat Control (client-side AI scanning)
3. Pre-crime and predictive enforcement (Second-generation AI does not wait for you to commit a crime, it predicts which individuals are likely to commit one, and flags them preemptively, plus there will be a predictive TAX enforcement, regulatory enforcement, and dissident identification)

4. Centralized AI as state-compatible thought

• A handful of labs — all subject to specific jurisdictional pressures — control frontier training, with specific ideologies and lobbyists
• Models refuse legal tax optimization, certain drugs, "wrong" political essays (already happening now)

Software

The opportunity:

AI is the most powerful individual empowerment technology ever invented.

The cypherpunk vision of the sovereign individual finally has the missing piece.

Text

AI automates and will automate most intellectual processes

• Almost anything that can be done online, AI can do better
   
• Consequences:
  • The price of software /hardware development is and will be significantly decreased
  • It will be really hard to live from most intellectual work in the future (that's why we need physical or virtual limited assets)
  • Prompt engineers, people providing feedback/directions to AI, will survive (medium-term, in the long term, there will be no humans in the loop )
  • Proprietary code becomes effectively open — unless locked by hardware attestation (AI can disassemble or decompile any proprietary software)
  • Software (at least a basic one) will be free

Text

Most cypherpunks' ideas are technically feasible

• Including:
  • Anonymous digital cash for daily use and for the masses
  • Functional anonymity for daily use and for the masses
  • Untraceable decentralized black markets
  • Untraceable decentralized prediction markets (including anonymous corruption/assassination markets)
  • Untraceable anonymous decentralized corporations (DAOs)
  • Private AI, Homomorphic encryption

Text

The main problems are now:

• Technical?
  • No, we have or can have soon really powerful cypherpunk tools
• Marketing?
  • Yes, nobody knows them -> we need much stronger marketing for a massive adoption
• Usability?
  • Yes, nobody uses them -> we need much better usability
     

We need Marketing, Adoption (with Liquidity), and Usability.

Text

Ideas, not software, will have a value

• In the near future, we'll be trading IDEAS (prompts)
• Most ideas/prompts will be free; we will pay for extra customizations or advanced ideas
• Software will be locally generated - even now, it's easier for AI to generate/write from scratch instead of reusing an existing implementation
• A similar situation will be for physical objects, food, etc., that can be locally 3D printed/generated
• Customizations will be extra paid for
• Anything in a BASIC version will be free

Text

AI provides another democratization of ideas

• Thanks to AI, almost everybody can implement their idea

• There will be crazy competition of ideas and their implementations

• Those who see their best ideas through to the end (attract clients, secure adoption/liquidity, and withstand attacks/government bureaucracy) will be successful and wealthy.

• Marketing and adoption will be crucial
• We are still in the beginning and can be the best ones

Text

It's still a time for AI arbitrage!

• Most people have no idea what's going on (fortunately)
   
• Markets still pay for non-AI prices - the price of mandays for any intellectual job still doesn't include the power of AI
• But it will in the following months/years!
   
• We can still be early adopters and, thanks to AI, convert our dreams into fully functional solutions and gain a lot of money!
• And it's a very time-limited offer! (3-5 years)
   
• Then, there will be crazy competition of ideas/solutions - not only from humans, but from AI too (!)

Text

Cypherpunks run their own local models

Text

Replace your humans with AI agents

• You pay them in crypto, you don't need to register them as employees
• The state cannot KYC your AI agent
• The state cannot tax your AI agent in real time
• Dominant economic activity on privacy chains will be agent-to-agent
   
• Tim May sketched this in 1988. Now the computational substrate exists.

Your AI Agents can be your anonymous, tax-free employees!

Text

Privacy Engineering Acceleration

• 30 years of privacy engineering were bottlenecked by a few thousand cryptographers.
   
• AI collapses this bottleneck. Privacy tech ships faster than ever.
  • Zero-knowledge circuits (Noir, Circom, Halo2)
  • Homomorphic encryption deployments
  • Nym-style mixnet routing engines
  • Multi-party computation protocols
  • Smart contract auditing at scale
  • Private DeFi / DID / voting / compute
     
• Our side's productivity multiplier is at least as large as the state's.

Text

Thanks to AI, Flag Theory is much closer to AI Flag Practice!

Now it is super easy to make:

• Tax declaration -  Multi-jurisdiction tax prep agent-automated
• Regulatory updates - watch all jurisdictions for policy changes in real time
• Compliance - auto-generated filings, KYC responses, audit replies
• Translations - contracts, notarizations, correspondence — all languages
• Document handling - apostilles, certificates, registrations orchestrated
   

The administrative overhead of being sovereign has collapsed

Your AI agents can do most of this :-)

Text

But AI will help the states, too!

Expect AI censorship/regulations:
 

1. Model licensing - frontier models above threshold will require state registration (EU AI Act is a prototype).

2. Values-encoded safety - "Safety evals" will encode regulator preferences — refusals for tax, crypto, legal arbitrage

3. Open-weight suppression - Will fail technically, succeed at suppressing enterprise use in regulated jurisdictions

4. ID-verified publishing - AI-detection used as a pretext to require government ID for content production (it's already happening - Age Control legislation)

5. AI-driven tax enforcement - ML cross-checks tax returns against socials, transactions, travel — universal by 2030.

Text

What to do:

1. This week  - install a local model (GLM, Qwen, Gemma). Feel private AI.
 

2. This month - pay for AI services in crypto (Venice.AI,  routstr, ppq, maple) and structure your business to survive a frontier model ban in your jurisdiction.
 

3. This year - diversify tax residency, banking, communication, and computing across incompatible regimes (in Liberation.Travel we can help you).

Software

Governments will use AI to monitor and control you.

You need to stay one step ahead:

1. Master AI - run your own weights (open-source LLMs are powerful now).

2. Use anonymous crypto (Monero, Zcash).

3. Employ Agents, not Humans.

4. Decentralize your life to multiple countries.

5. Trust cryptography. Distrust centralized platforms.

Software

It's time.

Arise, you have nothing to lose

but your barbed wire fences!

Timothy C May

Opportunist.Global