"Ship beats paper."— Doug
Addendum: 24+ innovations documented and filed
| Innovation | Status |
|---|---|
| Style-locked structural pruning | ✅ Published |
| ALDO protocol (dream logic orchestration) | ✅ Published |
| Combinatorial dream seed engine (7-bag, 311B+) | ✅ Published |
| The Pad — platform architecture | ✅ Published |
| Tadpole / Lucid dual-stream modes | ✅ Published |
| Touch-noise injection at coordinates | ✅ Published |
| img2img feedback loop for dream evolution | ✅ Published |
Effect: Prior art established. Any future patent applications covering these techniques can be challenged and invalidated.
The publications say what we do. The trade secrets are how we do it well.
| Component | Strategy |
|---|---|
| General pipeline architecture | Potentially open — published anyway |
| ALDO concept | Potentially open — reference implementation |
| Stylus training pipeline | Keep closed — trade secret |
| WebRTC streaming layer | Potentially open — commodity tech |
| Dream seed bag contents | Keep closed — trade secret |
Open what builds community. Close what builds competitive advantage.
Why this matters: Aligns artist incentives with the platform. Artists aren't giving away their style — they're monetizing it with full ownership.
| Factor | Our Position |
|---|---|
| Technical moat | 6+ months head start. Pruning recipes are trade secret. Execution details matter enormously. |
| Speed moat | We ship weekly. They ship quarterly. |
| Artist relationships | Artists own their Stylus and have reason to stay. Big corps would want to own everything. |
| Patent risk to us | Low — defensive publications establish prior art for all core innovations. |
| Patent risk from them | They can't patent our published techniques. They'd need genuinely novel improvements. |
| Acquisition play | Most likely outcome if we succeed. Clean IP story makes us an attractive target. |