A leave-behind for a conversation · unlisted · independent work
Anthropic openly studies how AI reshapes the labor market: the Economic Index, the $200M+ Economic Futures program, Dario's Machines of Loving Grace. The research asks where displaced and transitioning people go, and how to fund their transition. There is one thing the map still needs: a real, funded, proof-verified destination where a person does dignified work today.
EcoWealth runs that destination, the Ecological Work Protocol: real ecological, construction, food and care work as three competence ladders (Build · Grow & Cook · Restore), step-one-anyone-today to mastery, funded and proof-verified (photo + GPS + method), settled on Base mainnet with tCO2e receipts. This brief proposes plugging that destination into your research and your tools, so "AI will change work" becomes "here is where the humans go, and it's good."
✻ Said plainly, and warmly: EcoWealth is built on Claude. This very brief was written by Claude, running the engine. We're a genuine builder and daily user proposing a fit, not a vendor cold-calling. Not a teardown; a value exchange between a mission and the place that mission's people can land.What we found: Anthropic's own research maps where AI-displaced work goes, but names no funded, proof-verified destination where that work actually lands.
What we built, free: a real one: the Ecological Work Protocol, live on Base mainnet, plus a Claude-routed demo and an MCP agent kit, both reviewable below.
The ask: fifteen minutes (the same we'd give any partner), nothing requiring changes to your systems.
The value exchange
Three documents, one thesis
Anthropic's own published findings on AI and labor, cited to the exact URLs, then the gap (the research maps the disruption and funds the study of transition, but no dignified, funded, proof-verified destination is wired to it), where the Work Protocol plugs in, a light honest agent-readiness read of your public surface, and the ask. The heart of the case.
Read the brief →A brand-matched, clickable demo: a person says "AI is changing my field, I want real work that matters," and Claude routes them into a ladder, step one today, funded, proof-verified, settled to a receipt.
Watch it run →The actual files that let any assistant route a user into the Work Protocol: find_work, start_ladder, fund_or_credit, submit_proof, settle_and_receipt, spoken in MCP, your own standard.
Credit where it's due: your developer platform already ships a 200KB llms.txt (the file AI assistants read first). The economic-research and marketing surfaces don't yet. Passive, respectful, evidence-logged.
Grounded in Anthropic's own words: every position cited
AI's effects on labor markets and the economy over time, measuring real Claude usage across occupations, tracking 57% of tasks augmented, 43% automated, open-sourcing the data. anthropic.com/research/the-anthropic-economic-index
meaning comes mostly from human relationships and connection, not from economic labor, and in the long run our economic setup
will no longer make sense, needing a broader conversation on how it's organized. darioamodei.com/essay/machines-of-loving-grace
Transition support:
What mix of retraining, income support, and job matching would be most effective?, a $200M research fund and a $150M fellowship. anthropic.com/economic-futures/program
the long-term benefit of humanitythat strives
to communicate what we're seeing at the frontier to policymakers and civil society.A funded, dignified human-work destination is the civil-society other half. anthropic.com/company
The hand-off
Your research names transition support and job matching as the work ahead. EcoWealth is the destination that makes those words concrete: real ecological and community labor, funded, proof-verified, settled on-chain, reachable through Claude over MCP. Start with the brief; watch the demo run in a palette drawn from your own.
Try it in an afternoon: over MCP or plain HTTP
The destination isn't a slide, it's a live endpoint Claude can reach today, over MCP or plain HTTP. No account with us, no API key, no sales call. Here is the whole path.
Fund it a few cents. Give any AI agent (Claude, GPT-class, Llama, or your own) a Base wallet holding a few cents of USDC. That's the whole setup.
Point it at the endpoint. The work-packet program answers HTTP 402 with a standard x402 payment schema (a web-native way for AI agents to pay per use); the agent signs a USDC authorization and pays $0.10. A runnable client with a dry-run default is published at vealth.net/recipes.html.
Get a verifiable packet. Back comes a bounded, proof-verified ecological work packet, and every step the agent took (payment, settlement) is an on-chain receipt anyone can verify, forever. The EWP contract is live on Base mainnet; workId 14 settled the first packet end-to-end.
Why this isn't scary
The agent cannot do anything invisible here. Every action embeds its own receipt: permanent, public, verifiable. That is more accountability than most enterprise software offers, not less.
The bet
Your own Economic Index measures real Claude usage: 57% of tasks augmented, 43% automated, as AI competence scales. Human, proof-verified physical work an agent coordinates but cannot perform is a new augmentation signal you could count today. Being early costs $0.10 and one afternoon. Being late means the proof-verified track records belong to someone else.
The one concrete ask: have Claude, over MCP, make one $0.10 call and route a transitioning person into their first funded, proof-verified rung, nothing changes on your systems.
