LAGRANGE

LAGRANGE

52,483 Twitter followers

Follow

Hyper-parallel ZK coprocessing. Verifiably compute over blockchain data at a big data scale and unlock new data-intensive and interoperable applications.

Posts

Distributed autonomous systems introduce new failure modes. A single compromised node can influence swarm behavior. When systems act collectively, integrity of coordination becomes everything.

Checking in from @OneRSAC in San Francisco. Good to catch up with partners across the floor, from @awscloud and @IBM to @databricks. A few themes so far with AI showing up everywhere: • More attacks are relying on deception (deepfakes, identity spoofing) rather than breaking systems directly • Many teams are shifting from trying to prevent every breach to assuming compromise and limiting damage quickly Also, robots handing out security advice now, so the bar keeps moving.

Major technical milestone. We just achieved 60% performance improvement in our zkML proving pipeline. Accuracy now matching PyTorch float inference. Models like Gemma3 and GPT2 now prove faster while maintaining float-level precision. By merging quantization and lookup operations and refactoring non-linear layers, we accelerated proving without sacrificing correctness. Performance gains don’t matter if accuracy regresses. Now they don’t.

In mission-critical systems, uncertainty is risk. If autonomous agents don’t share a consistent mission state, coordination breaks down. And in distributed systems, small inconsistencies can cascade fast.

Checking in from @nvidia GTC in San Jose. A few themes showing up everywhere: • AI focus is shifting from model training to real-world deployment • Agentic systems are moving from answering questions to taking action • NVIDIA is expanding beyond chips toward the full stack - hardware, software, and infrastructure Seeing companies like @QuantinuumQC, @EvidenceOpen , and @Infleqtion pushing the boundaries of compute makes one thing clear: As AI moves into real products, the conversation is shifting from performance to reliability, control, and verification.

Cryptography turns “trust me” into “prove it.” That’s the upgrade defense AI can’t skip.

.@Consensus_HK Hong Kong wrapped last week. Over five days of panels, side events, and institutional gatherings, one theme kept surfacing in every room: AI is moving from demos to deployment, and deployment requires verifiability.

Checking in from Riyadh at World Defense Show This week’s conversations continue to center on what it takes to deploy AI into real defense environments and what’s required to make those systems verifiable and accountable. Let’s talk and build together: • Verifiable inference • Integrity of autonomous decision logs • Cryptographic guarantees for deployed AI • Accountability in mission-critical systems

Thanks for having us today. We couldn’t agree more. We’re building toward a verifiable, trusted future in the age of AI. twitter.com/lagrangedev/status...

Team Lagrange is headed to Hong Kong for Consensus. Our Head of BD, Brian Novell, will be speaking on verifiable AI, agent systems, infrastructure, and real-world deployment. Here’s where you can catch us - and register 👇

Loading..