The Brief
New briefs daily around 7 AM Eastern
Anthropic's FlashProve proves which model wrote a reply at 110 tokens/s
Until now a customer had to trust a router or provider that the model they paid for answered. A zero-knowledge proof makes that checkable without exposing weights, at roughly 70-95x the GPU cost of plain inference.
- 01Read
Postgres 19 slips as LLM-found bugs push back-patches up over 60%
A mature database is now finding old bugs faster than its committers can fix them, and the cost is a late major release with large features reverted. Projects adopting AI bug finding inherit that patching load.
- 02Read
MIT CSAIL shows GPT-6 Astra reading and editing neural network weights
Astra fixed Qwen3.5-0.8B reasoning loops by editing 0.016% of its weights, with no training run. Model repair has meant gradient updates and eval cycles; this is a direct edit, and the full experiment logs are public for others to check.
- 03Read
Claude Messages API adds server-side message threads in public beta
Clients send only new messages plus the previous message ID instead of the full history each turn. Long-running agents stop re-uploading growing transcripts, and Claude begins responding sooner because the server already holds the state.
- 04Read
OpenAI raises critical bug bounties fourfold to as much as $172,000
OpenAI is pairing the higher bounties with an internal sprint where Cathedral Systems researchers use its models against its own defenses, and with wider model access for trusted defenders. Frontier models are becoming a standard tool on the defensive side.
- 05Read
Datadog cuts Bits investigation agent cost from $25 to $4 per alert
Run time also fell from 11 minutes to about 2, and quality rose 44% on Datadog's internal benchmark. At $25 per alert, running an agent on every page was expensive; at about $4 the per-alert cost drops by more than 80%.
- 06Read
Stripe issues claimable sandboxes with test keys before any signup
An agent building a payment integration gets working test API keys with no registration, and a person claims the account later. Signup was the step that stopped an agent and waited for a human; Stripe built this with Vercel, Replit and Lovable.
A dated brief from the vibe-coding frontier. Today’s Intel.