Archive
136 articles, grouped by date.
zkML, optimistic, TEE, FHE all prove the computation. Restaking takes the other road: bond it and slash liars. We do the cost-of-corruption math behind EigenLayer's $18B AI-AVS security, the overloading attack that breaks it, and the probabilistic-audit tax.
Story raised $140M to make IP programmable for the AI era. We read its chain: 6.46M registered IP assets, 230,300 licenses ever minted, one week of royalties totaling $10 — plus the LAP/LRP royalty math and the precompile at 0x0101 that computes the graph.
Render just voted to route 60,000 Salad GPUs' payments through a token burn; Akash auctions its 234 GPUs per-block on-chain. Two opposed pricing mechanisms, live H100 quotes, and the burn-vs-emission math the equilibrium story skips.
An autonomous agent can't sign up for an API key or swipe a credit card. x402 revives HTTP 402 and settles in gasless USDC — here's the EIP-3009 handshake, the facilitator trust model, and where prompt injection breaks it.
Training Llama-3.1-8B costs ~$6M. The weights are free to clone. Perinucleus sampling embeds 24,576 distinct fingerprints per model with no quality loss — and a Poseidon ZK commitment registers ownership on-chain for $0.017, enforceable by smart contract.
Every zkSNARK bottoms out in a pairing check. Before Pectra you had BN254 at ~100-bit security. EIP-2537 added BLS12-381 at 128 bits, and its pairings are cheaper: 32,600k+37,700 vs 34,000k+45,000 gas. G2 MUL dropped from millions to 22,500 gas.
Inference markets charge a flat price per query. A cascade arbitrageur routes easy tasks to a cheap model and escalates only failures — capturing the spread. Olmedo, Schölkopf & Hardt (2026) show 40% net margins with no model-dev risk.
Majority voting counts hands — it can't rank quality. Fortytwo's Bradley-Terry protocol achieves 85.90% on GPQA Diamond vs. 68.69% for majority voting, extracting ranking signal from pairwise comparisons and making Sybil attacks economically unattractive.
Nova replaces expensive SNARK-in-SNARK recursion with a random linear combination that folds two R1CS instances into one — 10× less overhead per step. For N-step AI inference: O(N) prover work, one constant-size final proof. NANOZK delivers this at GPT-2 scale: 6.9 KB, 23 ms verify.
EIP-7702, live in Pectra, lets an EOA borrow smart contract code for one transaction — no migration, no bundler. For AI agents: atomic batching closes the approve-to-swap MEV window, and session key delegation bounds the blast radius of a compromised agent without touching the master key.
BERT-base under FHE takes 149 seconds, needs 112 GB, and sustains 0.007 samples/sec. Best-case MPC takes 2.3 seconds and 4 GB. The gap traces to one operation: bootstrapping — the noise-refresh that makes FHE theoretically unlimited but practically expensive.
Skill marketplaces let any agent buy trading capabilities off a shelf — and every skill is an execution vector. SAE's non-bypassable execution guards cut max drawdown from 46.4% to 3.2% (↓93%), CVaR by 97.5%, and the Delegation Gap by 97% on live perp data. The defense isn't in the model.