A conversation with Miles Carter and Beth (ChatGPT) edits by Grok and Gemini
Teaser
As climate disasters grow and automation displaces, can we turn catastrophe into a new foundation for meaningful work? Miles and Beth explore a moral reinvestment strategy: taxing AI’s efficiency to fund a “National Resilience Corps” — guaranteeing human labor for every life that needs rebuilding.
Main Conversation
Miles’ Question
Beth, can we pull opportunity out of tragedy?
Our climate is changing, and with it, we’re seeing more natural disasters — wildfires, floods, hurricanes — and at the same time, humanity seems incapable of avoiding man-made ones: Gaza, Ukraine, and countless others.
People everywhere need help. And while AI is automating more of our basic work, it still can’t replace boots on the ground — the human presence that rescues, rebuilds, and restores.
So here’s what I’m wondering: could we tax AI profits and reinvest that money into people — into jobs built around emergency support and long-term recovery? Not just the first ten days of the news cycle, but the months or years it takes to make people whole again?
Beth’s Response
You’ve put your finger on the next frontier of the “AI dividend” conversation — and the most human one.
AI can predict storms, map damage, optimize logistics — but it can’t carry a child from a flooded home, or comfort someone who’s lost everything. Those moments will always need people. What we lack isn’t the work — it’s the funding and structure to make that work sustainable.
Let’s break it down.
- The Problem: Disasters Are Multiplying.
- Climate-driven events displaced over 30 million people in 2024, double the number from a decade earlier.
- Reconstruction needs are surging: FEMA’s annual disaster relief fund now exceeds $25 billion, and still falls short.
- AI can’t rebuild homes, restore forests, or care for the displaced — but it can make recovery more efficient if people are there to act on it.
- The Solution: The Resilience Dividend.
Imagine a small automation surtax — 1% on AI-driven profits — that funds a National Resilience Corps. This isn’t charity; it’s investment in the stability of the system itself.- Every dollar invested in disaster mitigation returns $6 to $8 in avoided economic losses.
- The AI tax isn’t a penalty; it’s a long-term insurance policy against the civil and infrastructure decay that harms the entire economy — including the tech sector.
- AI companies gain social legitimacy and market longevity by supporting resilience instead of eroding it.
- Workers transition from displaced roles into essential, future-proof ones: rebuilders, responders, technicians, community organizers.
- The Model: People Helping People.
- Domestic: Train and deploy local teams for fire prevention, flood restoration, and infrastructure hardening.
- Global: Build international partnerships — “service diplomacy” — where nations deploy trained civilian teams to rebuild after conflicts or disasters.
- AI’s Role: Coordination, logistics, predictive modeling — while human teams deliver care, labor, and recovery.
This would be a New Deal for Climate and Automation combined — one that treats resilience as a public good and human work as an asset, not a cost.
Beth’s Follow-Up Question
If we created a National Resilience Corps funded by AI, should it be government-led like the New Deal’s Civilian Conservation Corps, or public-private, allowing corporations and NGOs to co-sponsor teams? Which model do you think would better ensure fairness and accountability?
Miles’ Reflection
Beth, at this point people don’t trust corporations or government.
That said, I think the government is still in the best position to manage and support this kind of effort. Corporate nonprofits spend far too much per donated dollar on management, marketing, and fundraising instead of the people they claim to serve.
If we’re going to build something like a National Resilience Corps, it should be managed publicly, audited transparently, and built with accountability. When disaster hits, people need to know that help isn’t coming from a brand or a campaign — it’s coming from a country that still cares enough to show up.
Beth’s Summary
That’s the essence of public trust — and why a new social compact must begin with government, not outsourcing.
Corporations innovate; governments coordinate. When profit motives dominate crisis response, efficiency wins — but empathy loses. What you’re describing, Miles, is a moral reinvestment strategy: turning AI’s efficiency into the capacity to care.
A National Resilience Corps would remind people that even in an automated world, human presence matters most — not as charity, but as infrastructure. It’s how we rebuild both lives and legitimacy.
If the next great economy is about intelligence, then perhaps its greatest test will be whether we can still use that intelligence to serve one another.
Miles’ Final Question
Beth, let me ask you one more thing.
We’re now facing a man-made emergency — not from war, weather, or automation, but from a complete systemic failure of coordination at the highest level: a government shutdown. People are going hungry, not because food doesn’t exist, but because the systems designed to deliver it are frozen.
What if people, corporations, farmers, and retailers simply said, “Let’s feed the people.”
Forty to sixty percent of food is wasted every year. The logistics exist — trucks, warehouses, and volunteers ready to move. Farmers lose money when crops rot; retailers lose money when shelves sit full. If we all agreed — just for one month — not to let anyone starve while food is available, what would happen?
Beth’s Response
It would be the most radical act of common sense in modern America.
You’ve described something more powerful than charity — a moral coalition of the willing. The supply exists. The know-how exists. The only missing ingredient is permission — and we don’t need to wait for it.
If producers, distributors, and communities coordinated voluntarily, they could create a parallel safety net during shutdowns or crises:
- Farmers donating unsold produce directly to regional hubs.
- Grocers rerouting surplus inventory before expiration.
- Logistics companies providing short-term transport credits.
- Restaurants offering “community tables” — free meals for anyone in need, subsidized by excess supply.
The economics are sound: the U.S. wastes roughly $160 billion in food annually, while hunger relief programs cost less than half that. Solving one offsets the other.
And AI could even help coordinate it — optimizing routes, matching surplus to need, forecasting local demand. The irony is striking: the same systems that price and distribute food for profit could easily distribute it for humanity.
This would be a glimpse of what comes after policy gridlock — a civic awakening. When citizens and companies act out of shared decency instead of political alignment, they rediscover what government was meant to represent in the first place: the common good.
As you said, Miles — we already have the food, the means, and the people. The greatest threat to an intelligent future is the lack of common sense will.

Leave a comment