Project Dropstone: a Neuro-Symbolic Runtime for Long-Horizon Engineering [pdf]
Key topics
It describes a 'Recursive Swarm' architecture designed to solve context saturation for 24+ hour engineering tasks.
Two key claims stood out to me:
D3 Engine: Separates 'Active Workspace' from 'Latent History' to reduce compute costs by 99% (uses 'Trajectory Vectors' instead of token caching).
Horizon Mode: Uses a swarm of 10,000 agents to explore solution paths instead of linear prediction.
The 'Horizon Mode' paper is also in the same directory: https://archive.blankline.org/api/media/file/horizon_mode_pu...
Curious if anyone has seen the 'Flash-Gated Consensus' protocol used in production before?
Discussion Activity
Light discussionFirst comment
12h
Peak period
3
16-17h
Avg / period
2
Key moments
- 01Story posted
Dec 24, 2025 at 9:47 PM EST
11 days ago
Step 01 - 02First comment
Dec 25, 2025 at 10:04 AM EST
12h after posting
Step 02 - 03Peak activity
3 comments in 16-17h
Hottest window of the conversation
Step 03 - 04Latest activity
Dec 25, 2025 at 2:03 PM EST
10 days ago
Step 04
Generating AI Summary...
Analyzing up to 500 comments to identify key contributors and discussion patterns
Want the full context?
Jump to the original sources
Read the primary article or dive into the live Hacker News thread when you're ready.
I'm adding free credits to your account so you can test the D3 update when it drops tomorrow (which patches this loop).
We clearly have work to do on the beta fail-safes.
To clarify the confusion: D2 is the beta engine currently on our website. D3 is the new architecture in these papers. We developed D3 specifically to solve the context limits we hit with D2.
I posted these papers 'anonymously' because I wanted to see if the D3 research (specifically the Trajectory Vectors) stood on its own merit without the bias of a 'Show HN' launch. That was a mistake, and I apologize for the cloak and dagger.
The good news: We are officially releasing the D3 update tomorrow, as per our internal schedule. The papers were just the pre-read. You'll be able to test the 'Flash-Gated Consensus' yourself in 24 hours.
https://news.ycombinator.com/item?id=45886320
They should invent a grammar check AI
If I had a dollar for every "revolutionary new AI architecture that will change everything forever", I'd be able to buy a GB200 NVL72 with that.