Get your free personalized podcast brief

We scan new podcasts and send you the top 5 insights daily.

Tom Milton says a network certifier is an FPGA that stores a cryptographic log of traffic, enabling zero-knowledge proof that a cluster only serves inference. But FPGAs must be designed, security tested, red teamed, manufactured, and trusted by all parties. If a deal arrived next month, the hardware would not be ready.

Related Insights

Even if a startup creates perfect AI verification technology, it can't be used in a crisis until it's vetted by the intelligence community to become 'national technical means'—a process that historically takes years. Startups must engage with government agencies early to pre-vet their tech.

NeoClouds that use non-standard networking hardware instead of the NVIDIA default face a crucial trade-off. New open-source AI software is almost exclusively developed and optimized for NVIDIA's network. This means customers on custom networks can wait months for support for new features, a costly delay in the fast-moving AI space.

Tom Milton describes Emoto's prototype on about 32 H100/H200 GPUs: a network tap copies traffic to a verifier node, which reruns the claimed inference and checks outputs match. Unexplained packets reveal hidden workloads like training. This avoids inspecting GPU instructions directly.

Tom Milton argues verification is not only for outside regulators. If AI models gain nation-state-level cyber capability, labs need hardware-backed logs to confirm models have not changed their own weights or spoofed evaluation results. The same stack protects against untrusted AI as against an adversary country.

Tom Milton says the fallback is turning data centers off, but that is extremely costly. Verification lets a prover show it is serving only inference, or running small safety experiments, so a slowdown can preserve useful work while stopping dangerous training. This reduces the cost of reaching a deal.

Technologies like Intel TDX and NVIDIA's Confidential Compute encrypt AI workloads directly on hardware. This guarantees that even the physical server owner cannot access the data, allowing anyone to contribute hardware to a decentralized network without needing to be vetted or trusted.

Instead of relying on legal service level agreements (SLAs), Bittensor subnets use validators that programmatically check miner performance. They can use ZK proofs to verify an AI model's output, ensuring quality and honesty without trust.

Public announcements for massive new data centers may be "pollyannish." The reality is constrained by long lead times for critical hardware components like power generators (24 months) and transformers. This supply chain friction could significantly delay or derail ambitious AI infrastructure projects, regardless of stated demand.

Tom Milton says a government could require a zero-knowledge proof that a safety classifier is running on each API response. Then it could send random requests and check the proof remains attached, without inspecting the lab's hardware. This gives a correctness claim, though not proof no illegal model runs elsewhere.

Tom Milton says recomputation requires the verifier to access model weights inside the prover's data center, which labs dislike. Zero-knowledge proofs would let provers demonstrate correct inference without sharing weights, though current ZKML is orders of magnitude slower; sampling one in a thousand packets could suffice.