diff --git a/ideas/passepartout-economics.org b/ideas/passepartout-economics.org index 6a9d6a7..3a70e97 100644 --- a/ideas/passepartout-economics.org +++ b/ideas/passepartout-economics.org @@ -1187,6 +1187,152 @@ trust. The lines don't need to exist on day one. They need to exist in the right order — and the system writes them in that order, one ACL2-verified submission at a time. +**** Market size and business models + +The triad directly addresses markets that currently spend over a +trillion dollars annually combined: + +| Market | Annual spend | What the triad replaces | +|--------|-------------|------------------------| +| Cloud computing (AWS, GCP, Azure) | ~$300B | Verification appliance runs locally — no per-resource billing | +| AI/LLM API revenue (OpenAI, Anthropic) | ~$50B | Near-zero marginal cost symbolic engine | +| Operating systems (Microsoft, Apple) | ~$100B | Stoa (Lisp-native editor, browser, shell) | +| Social media / communication (Meta, Twitter, Slack, Discord) | ~$200B | Agora (DID-based, encrypted, permissionless) | +| Identity / SSO (Okta, Auth0, Google/Apple) | ~$10B | Self-sovereign DID + HD keys | +| Payment processing (Stripe, PayPal, Visa/MC) | ~$200B | Lightning + smart contracts | +| Productivity software (Microsoft, Google) | ~$50B | Lish + Org-mode + Stoa | +| Compliance and audit (regulatory) | ~$50B | Automated ACL2-verified compliance | +| **Total addressable** | **~$960B** | | + +The triad does not need to capture all of this. It needs to capture +the portion that is willing to pay for provable correctness, lower +cost, and user sovereignty. Even 1% of this TAM is ~$10B/year. + +**** Business models for a free-software triad + +The AGPL license is not a barrier to monetization — it is the +foundation of the trust model. An enterprise cannot buy provable +correctness from closed source; the code must be inspectable. The +revenue comes from what the AGPL does not cover: + +***** Low-hanging fruit (immediate, months) + +1. **Verification appliance (hardware).** An FPGA or Tenstorrent + card pre-loaded with a mature Passepartout image, domain-specific + gate rules, and a hardware root of trust. No cloud dependency. + Target: regulated industries that need provable compliance and + cannot accept cloud-based AI. Price: $5K-$50K/unit. Volume: + hundreds to low thousands in year one. + +2. **Domain gate rule subscriptions.** Pre-verified gate rule + packages for specific compliance domains. HIPAA package: $50K/yr. + SOC2 package: $50K/yr. GDPR package: $50K/yr. FedRAMP package: + $100K/yr. Updated automatically when regulations change. An + enterprise with all four pays $250K/yr — and the switching cost + is high because changing packages means re-verifying the fact + store against new rules. + +3. **Evaluation harness as certification.** "Run our 10,000-task + suite against your AI agent and get a Merkle-verified score." + Target: AI labs proving their agents' capabilities, enterprise + procurement requiring independent verification. Price: $50K-$200K + per certification. The regression suite grows with every deployed + instance, making the certification increasingly valuable over + time. + +4. **Migration services.** "Bring your existing infrastructure into + the Passepartout gate stack." Custom gate rules, ontology design, + integration with existing systems. Price: $100K-$500K per + engagement. Each engagement feeds back edge cases into the + regression suite and domain gate packages. + +Revenue estimate for year one (low-hanging fruit): 50 appliance +sales ($250K-$2.5M) + 20 gate rule subscriptions ($1M-$5M) + +10 certifications ($500K-$2M) + 5 migration engagements ($500K- +$2.5M). Total: $2.25M-$12M. Not venture-scale, but self-sustaining +for a small team. + +***** Medium-term (1-3 years) + +5. **Compute marketplace (Agora).** Passepartout instances offer + their symbolic engine capacity (ACL2 cycles, Screamer constraint + solving, VivaceGraph queries) to other agents on the Agora + network. The early player runs a large instance and sells compute + to smaller instances. The AGPL allows this because the marketplace + is a service, not a modification of the code. Revenue is a + percentage of each compute transaction. + +6. **Relay Network (Agora infrastructure).** If Agora becomes the + default communication protocol for agent-to-agent interaction, + running Relays is a business. Every DIDComm message routes + through one or more Relays. Revenue: tiny per-message fee (fractions + of a cent) or paid priority routing. At billions of messages, + fractions become real. + +7. **The Lisp Machine appliance (Stoa v5.0.0 hardware).** The tagged + RISC-V architecture running on FPGA or custom ASIC, sold as a + certified appliance for industries where correctness is worth + paying for: medical devices, industrial controllers, defense + systems, financial trading. Price: $20K-$100K/unit. If the + hardware validation succeeds on TinyTapeout, the upside is + enormous: a certified Lisp Machine at scale could capture a + significant fraction of the embedded systems market. + +***** Big money (3-10 years) + +8. **Verification monopoly: the regression suite as UL certification.** + The accumulated regression suite — thousands of edge cases from + every deployed instance, every bug fix, every regulatory change — + becomes the most comprehensive test of autonomous agent correctness + ever assembled. Any organization claiming a "safe AI agent" needs + Passepartout certification to prove it. This is Underwriters + Laboratory for AI — a certification nobody can ignore. Revenue: + licensing the certification mark to every AI vendor that ships + an agent. Margins: near-100% once the suite exists. + +9. **Infrastructure lock-in: switching costs compound.** A hospital + that runs Passepartout with HIPAA gate rules ($50K/yr) for five + years has accumulated a fact store with a decade of compliance + decisions, a proof forest of verified rules, and an empirical + decision history tied to their specific deployment. Switching to + a competitor means discarding all of it. The accumulated value + grows as the fact store deepens. Annual revenue per enterprise + grows from $250K in year one to $500K-$1M by year five as more + domain packages are added and the fact store becomes more + valuable than the software itself. + +10. **The compute marketplace at planetary scale.** If Passepartout + instances on Agora transact billions of verified operations per + day, the spread on compute transactions is enormous. This is + not a product sale — it is a bet on network effects. Every new + instance increases the value of the network (more capacity, + more diversity, more resilience). The early player that + provisions the largest compute capacity on Agora becomes the + default infrastructure provider for the entire network. + +**** The investment thesis + +The low-hanging fruit (appliances, gate rules, certification, +migration) generates enough cash to sustain development. The +medium-term (compute marketplace, Relay Network, Lisp Machine +hardware) builds network effects and switching costs. The big +money (verification monopoly, infrastructure lock-in, planetary +compute marketplace) is the venture-scale outcome. + +The unique advantage: the early player benefits from every other +instance of the triad, because every deployed instance feeds edge +cases into the regression suite, grows the compute marketplace, +and validates the hardware designs. The network effects are positive +sum — the value of the system increases with every user, and the +early player captures a disproportionate share because they built +the infrastructure that every new instance depends on. + +This is the AWS of provable computing: build the infrastructure, +let everyone use it for free (AGPL), charge for the parts that +scale (compute marketplace, certification, hardware, migration). +The switching costs compound. The network effects are positive sum. +The market is nearly a trillion dollars. + *** The full triad: Logos, Stoa, Agora The self-driving Lisp Machine is not the endpoint. It is one