Endangered Developers Foundation
Alignment has a human side.
Whether AI goes well is not only a property of how models behave. It is a property of whether the economy AI produces stays aligned with the people inside it.
That alignment will not happen by default. It has to be engineered.
The pattern of progress, and where it stops
Every prior industrial transition displaced people, and every one eventually reabsorbed them, because the economy still needed human labour. Displaced weavers became factory hands; displaced clerks became knowledge workers. The bridge was painful, but there was always a far bank.
Artificial general intelligence changes the shape of the problem. If AI systems can do most cognitive work, and increasingly most physical work, then the economy's need for people is no longer guaranteed. The far bank may not exist unless someone builds it. Luke Drago and Rudolf Laine call this the intelligence curse: when value comes from AI rather than from people, the powerful lose their incentive to invest in people, the way resource-rich states neglect citizens whose labour they do not need.
The First Industrial Revolution
Steam power and mechanization displaced millions of agricultural and craft workers. Entire communities built around hand-weaving and manual labor were upended overnight. But within a generation, factory jobs, engineering roles, and new industries emerged — creating more prosperity than what was lost.
The Second Industrial Revolution
Electricity, steel, and mass production eliminated countless jobs in cottage industries. But they created the modern middle class — assembly lines needed supervisors, electrical grids needed engineers, and new consumer markets needed salespeople, marketers, and managers.
The Digital Revolution
Computers displaced typists, filing clerks, and manual record-keepers. But they created the software industry, the internet economy, and millions of knowledge-worker jobs that didn't exist a decade prior. Every wave of displacement was followed by a wave of creation.
The AI Revolution
Generative AI and large language models are automating cognitive work — coding, writing, analysis, design. What makes this one different is not the speed. It is that the reabsorption every previous wave relied on assumed employers would still need people. If AI can do most cognitive work, and increasingly most physical work, that assumption is the thing in question.
Why EDF exists
Developers are the first population fully exposed to this. They built the tools, they are being displaced by them first, and the pattern is already visible in the numbers: entry-level hiring collapsing in India and the US, layoffs attributed to AI rising from 7% to about 40% of all cuts through 2026, and nearly half in the first quarter.
An endangered species is a warning about the ecosystem, not a verdict on the species. What happens to developers now is the leading indicator for everyone who comes after.
Our thesis
Whether AI goes well is not only a property of how models behave. It is a property of whether the economy AI produces stays aligned with the people inside it.
That alignment will not happen by default. It has to be engineered: measured, funded, and built, starting with the population that needs it now.
We are not anti-AI. We are pro-human. The question is not whether AI will transform work; it already is. The question is whether the transformation keeps people in the economy or quietly writes them out of it.
Four pillars
Measure
Build the open Displacement Data Standard: an independent, methodology-attached record of who AI displaced, when, and what happened next. Every transition-insurance scheme, every AI dividend, every retention incentive needs a trigger. We are building the trigger.
Diffuse
Treat access to AI capability as a floor. Every developer in an EDF program will receive guaranteed frontier model access and compute for the length of their transition, and we advocate for that floor to extend beyond our cohorts.
Contribute
Turn the displaced into the people who make AI safer. The Alignment Corps trains developers for evaluations, red-teaming, and oversight tooling, and will place them with labs, safety organisations, and audit bodies as placement partnerships are signed.
Insure
Fund transition from the gains that cause it. We take a position of principle, not instrument: the productivity AI creates should pay for the transition it forces. Others have proposed token taxes, equity participation, wage insurance. We will pilot a private transition-insurance pool with employer partners and publish what we learn.
Three ideas we keep coming back to
Not replacement, and not elevation either
The comfortable story is that automation pushes people into higher-skilled, creative roles. It has been true so far. It is not guaranteed to stay true, and an organisation built on the assumption that it will would be building on sand. We plan for the harder case.
The gap is the test
Transitions harm people who lack bridges: training, funding, guidance, and now capability access. But the gap is also where a society reveals whether it still values its people once it no longer needs their labour.
From consumers to creators to contributors
Reskilling and incubation get people back to income. The Alignment Corps gets them a seat in the room where AI is made safer. That is the difference between being a beneficiary of the transition and being an author of it.
Our 2027 Targets
These are goals we are working toward, not results we have achieved. For what EDF has actually done so far, read where we are today.
10,000
Target: developers reskilled by 2027
500
Target: startups incubated
$5M
Target: micro-grants distributed
15
Target: policy papers published
50
Target: corporate & government partners
5,000+
Target: Displacement Data Standard v1 published with verified records
Supporting Research
Our mission is grounded in data. These reports from leading institutions document the scale and urgency of AI workforce displacement.
Reading that shaped our thesis
- Drago & Laine — The Intelligence Curse (essay series)
- Anthropic — Anthropic Economic Index, and the 81,000-worker survey
- Amodei — Policy on the AI Exponential
- World Economic Forum — Future of Jobs Report 2025
The future is being written now