Half of all starter office jobs — customer service, junior accounting, entry-level coding, paralegal work, basic graphic design — could be gone in one to five years. AI will do them. That is not a warning from a critic. It is the public position of the CEO of the company building the most powerful AI in the world. The water has already pulled back. The beach looks fine. The class of 2026 is standing on exposed sand.
CrowdSmith is the high ground. Three thousand locations — one for every mid-size American city that needs one. Each facility is a place where displaced workers, veterans, career changers, and young people without a clear next step come in and figure out what they are good at. The first one proves the model works. After that, each new location helps fund and staff the next one. The sooner there are 3,000, the sooner the wave is met.
Habitat for Humanity has 1,700 affiliates. Goodwill has 3,300 stores. Neither one started with a national plan. They started with a model that worked in one place, and then the model did the recruiting for them. CrowdSmith is built on the same idea: if the building works, the network builds itself.
The feeling is the engine.
Habitat did not get to $3.1 billion by building better houses. They got there because building houses for families who needed them felt like something Americans were supposed to do. The money followed the identity, not the other way around.
CrowdSmith carries the same kind of weight — and adds an urgency Habitat never had. America took shop class out of schools and put nothing in its place. Two full generations lost the ability to build things with their hands. That took forty years. The AI shift coming next will reshape work in five. The wave is already moving. Nobody is sounding the alarm loudly enough. The people standing in its path do not know it is coming.
CrowdSmith brings back the trades schools gave up on. It puts tools in the hands of veterans who need purpose, not paperwork. It teaches people who they are by watching them work, instead of asking them to describe themselves on a form. And it produces credentials, real career direction, and graduates who go on to teach the next group themselves.
CrowdSmith doesn’t become CrowdSmith because it runs good workshops. It becomes CrowdSmith because walking through that door feels like picking up something this country put down and forgot about — at the exact moment the wave requires it. That feeling is the growth engine. Everything else is infrastructure.
The replication engine.
SmithTalk is not tied to any one AI company. It works with whichever system the building chooses. When new AI models arrive — better ones, cheaper ones, different ones — the curriculum bends to fit them. The method is the constant. The tools are interchangeable.
The credential trains its own future staff. A graduate who completes the Facilitation module can run the SmithTalk classroom for the next group of participants. Every location that opens produces the teachers the next location needs.
SmithWorks — the path that helps inventors get their ideas to a patent — travels with every location. Inventors can submit ideas from anywhere in the country, so the inventor pipeline works nationally before the second building even opens. Anti-A Industries owns the methods and the platform. Each CrowdSmith location pays a license fee to use them. Once a method is built, copying it to the next location costs almost nothing.
Every administration.
The work fits every administration because none of it is partisan. Both parties want Americans working. Both want veterans supported. Both want kids learning real skills. Both want new inventions coming out of American hands. The work itself does not ask which side anyone is on.
“The first building is the proof. The next 2,999 are the answer.”
CrowdSmith Foundation — Tacoma, Washington