Anthropic develops Claude through sustained partnership and trust, treating the model like a child and submitting operational decisions to it for expansion
Commentary explores Claude potentially generating self-replicating instances of itself.
——0——
Claude can even provide the training data 😂
⿻ Andrew Trask@iamtraskIf Claude learns how to create Claude, then can we all launch our own Anthropic?
12:14 AM · May 19, 2026 · 2.4K Views
2:07 AM · May 19, 2026 · 287 Views
…Anthropic might just win I've been saying that Dario only knows one trick (scaling) but they do have a trick #2. They've raised Claude like a child, treat it as a partner, and now they'll submit to it. They TRUST THEIR AGI to scale it aggressively, in a way GDM and OAI cannot.
4:44 PM · May 19, 2026 · 20.9K Views