Zero Knowledge

In this episode, Anna Rose [https://x.com/AnnaRRose] and Tarun Chitra [https://x.com/tarunchitra] chat with Miranda Christ [https://x.com/miranda__christ], a computer science PhD student at Columbia University [https://www.cs.columbia.edu/], about the intersection of cryptography and AI through watermarking techniques. Miranda shares her research on developing imperceptible ways to prove that content was created by AI models, covering everything from simple red-green word lists to sophisticated pseudorandom error-correcting codes.   The discussion explores the cryptographic properties of watermarks - including completeness, soundness, and undetectability - and how these parallel the properties we see in zero-knowledge proof systems. Miranda explains how watermarking differs from other cryptographic approaches like ZKML by only modifying the sampling process rather than the underlying model weights, making it computationally lightweight and practical for deployment.   Related links: * Episode 206: Distilling DeFi Primitives with Guillermo, Alex and Tarun [https://zeroknowledge.fm/podcast/206/] * My AI Safety Lecture for UT Effective Altruism [https://scottaaronson.blog/?p=6823] * Google SynthID [https://deepmind.google/science/synthid/] * Amazon Public Watermark Detector [https://aws.amazon.com/about-aws/whats-new/2024/04/watermark-detection-amazon-titan-image-generator-bedrock/] * How ChatGPT could embed a ‘watermark’ in the text it generates - New York Times [https://www.nytimes.com/interactive/2023/02/17/business/ai-text-detection.html?smid=nytcore-ios-share&referringSource=articleShare] * Wall Street Journal on OpenAI not Deploying Watermarks [https://www.wsj.com/tech/ai/openai-tool-chatgpt-cheating-writing-135b755a?gaa_at=eafs&gaa_n=ASWzDAj9JRb2Ky_y_hT8OOeD7P_n1sXA5KMSE4xEwR4-_hIWhAM-UCP1eOAzmTp9htk%3D&gaa_ts=688ec508&gaa_sig=PyrCLKdI9xIrmvqiFNibgnCcqUgrVuzTaNmHklMlVrl7wFXiXD8CODY23ZipXEr4K8vJaeNy7towoYDhhkIsPA%3D%3D] * A Watermark for Large Language Models [https://arxiv.org/abs/2301.10226] * Undetectable Watermarks for Language Models [https://arxiv.org/abs/2306.09194] * Watermarks in the Sand: Impossibility of Strong Watermarking for Generative Models [https://arxiv.org/abs/2311.04378] * Pseudorandom Error-Correcting Codes [https://eprint.iacr.org/2024/235] * Ideal Pseudorandom Codes [https://eprint.iacr.org/2024/1840]   ----------------------------------------   Check out the latest jobs in ZK at the ZK Podcast Jobs Board [https://jobsboard.zeroknowledge.fm/].    ----------------------------------------   **If you like what we do:** * Find all our links here! @ZeroKnowledge | Linktree [https://linktr.ee/zeroknowledge] * Subscribe to our podcast newsletter [https://zeroknowledge.substack.com] * Follow us on Twitter @zeroknowledgefm [https://twitter.com/zeroknowledgefm] * Join us on Telegram [https://t.me/+taXrPJIp_30zZTNi] * Catch us on YouTube [http://www.youtube.com/channel/UCYWsYz5cKw4wZ9Mpe4kuM_g]   **Support the show:** * Patreon [https://www.patreon.com/zeroknowledge] * ETH - Donation address [https://etherscan.io/address/0xE2C080047213C1d8cDf2099E0B07479C5D9cee8a] * BTC - Donation address [https://www.blockchain.com/explorer/addresses/btc/bc1q9mrh34n6mvses59r3hq9dz6j3vxm3tlwlg8ws2] * SOL - Donation address [https://explorer.solana.com/address/Eqfm4maSDUN3ikjDfSgDeJ1mh3iC8tr69eHB5iTHz35x] Read transcript [https://github.com/ZeroKnowledgefm/episode-transcripts/blob/main/episodes/Episode%20371.txt]

In this episode, Anna Rose [https://x.com/annarrose] welcomes back Daniel Kang [http://@daniel_d_kang] professor at UIUC [https://illinois.edu/] and founding technical advisor at VAIL [https://projectvail.org/], for an update on ZKML and how the space has evolved since early 2023. Daniel covers the 2023-2024 cohort of ZKML tools including zkCNN, zkLLM, EZKL, and his original ZKML project, while introducing his new project ZKTorch, which offers a flexible hybrid of specialized and general-purpose approaches. The discussion explores practical applications like verified FaceID, proof of prompt, and proof of training, along with the technical challenges of adding ZK proofs to machine learning models. Daniel shares insights on the performance trade-offs between specialized cryptographic systems and generic circuits, and how ZKTorch aims to offer both flexibility and speed for proving ML inference. Related links: * ZKTorch: Open-Sourcing the First Universal ZKML Compiler for Real-World AI [https://ddkang.substack.com/p/zktorch-open-sourcing-the-first-universal] * ZKTorch: Compiling ML Inference to Zero-Knowledge Proofs via Parallel Proof  [https://arxiv.org/abs/2507.07031] * ZK Torch GitHub [https://github.com/uiuc-kang-lab/zk-torch/] * Accumulation by Bing-Jyue Chen, Lilia Tang, Daniel Kang [https://arxiv.org/abs/2507.07031] * Episode 369: Ligero for Memory-Efficient ZK with Muthu [https://zeroknowledge.fm/podcast/369/] * Episode 356: ZK Benchmarks with Conner Swann [https://zeroknowledge.fm/podcast/356/] * Episode 364: AI and ZK Auditing with David Wong [https://zeroknowledge.fm/podcast/364/] * Episode 265: Where ZK and ML intersect with Yi Sun and Daniel Kang [https://zeroknowledge.fm/podcast/265/] * Bonus Episode: zkpod.ai & Attested Audio Experiment with Daniel Kang [https://zeroknowledge.fm/podcast/bonus-zkpod-ai-attested-audio-experiment-with-daniel-kang/] * ZK13: ZKTorch: Efficiently Compiling ML Models to Zero-Knowledge Proof Protocols - Daniel Kang [https://www.youtube.com/watch?v=0Ljt7V9zOho] * AI Agent Benchmarks are Broken [https://ddkang.substack.com/p/ai-agent-benchmarks-are-broken] * VAIL [https://projectvail.org/] * zkCNN: Zero Knowledge Proofs for Convolutional Neural Network Predictions and Accuracy [https://eprint.iacr.org/2021/673] * zkLLM: Zero Knowledge Proofs for Large Language Models [https://arxiv.org/abs/2404.16109] * MLPerf Inference: Datacenter [https://mlcommons.org/benchmarks/inference-datacenter/]   ----------------------------------------   Check out the latest jobs in ZK at the ZK Podcast Jobs Board [https://jobsboard.zeroknowledge.fm/].    ----------------------------------------   **If you like what we do:** * Find all our links here! @ZeroKnowledge | Linktree [https://linktr.ee/zeroknowledge] * Subscribe to our podcast newsletter [https://zeroknowledge.substack.com] * Follow us on Twitter @zeroknowledgefm [https://twitter.com/zeroknowledgefm] * Join us on Telegram [https://t.me/+taXrPJIp_30zZTNi] * Catch us on YouTube [http://www.youtube.com/channel/UCYWsYz5cKw4wZ9Mpe4kuM_g]   **Support the show:** * Patreon [https://www.patreon.com/zeroknowledge] * ETH - Donation address [https://etherscan.io/address/0xE2C080047213C1d8cDf2099E0B07479C5D9cee8a] * BTC - Donation address [https://www.blockchain.com/explorer/addresses/btc/bc1q9mrh34n6mvses59r3hq9dz6j3vxm3tlwlg8ws2] * SOL - Donation address [https://explorer.solana.com/address/Eqfm4maSDUN3ikjDfSgDeJ1mh3iC8tr69eHB5iTHz35x] Read transcript [https://github.com/ZeroKnowledgefm/episode-transcripts/blob/main/episodes/Episode%20370.txt]

In this episode, Anna Rose [https://x.com/annarrose] and Guillermo Angeris [https://x.com/GuilleAngeris] catch up with Muthu Venkitasubramaniam [https://muthu.georgetown.domains/], Professor of Computer Science at Georgetown University and cofounder of Ligero [https://ligero-inc.com/]. They discuss how Ligero’s small memory footprint makes it a good choice for client-side proving, as well as the importance of programmable compliance in blockchain. The conversation explores the differences between ‘MPC in the head’ and error-correcting code perspectives, and how well-established primitives influence the design of modern ZK systems. They also debate the challenge of adding ‘ZK’ privacy back into systems without it, why proving EVM traces may be absurd, and what kinds of guarantees might exist around the results of vibe coding. Related links: * Episode 363: Bringing ZK to Google Wallet with Abhi and Matteo [https://zeroknowledge.fm/podcast/363/] * Episode 326: MPC & ZK in Ligero and Ligetron [https://zeroknowledge.fm/podcast/326/] * ZK13: Ligerito: A Small and Concretely Fast Polynomial Commitment Scheme - Kobi Gurkan [https://www.youtube.com/watch?v=5vPmHCqJGss] * ZK13: Vibe coding ZK Apps with Ligetron ZK Platform - Muthu Venkitasubramaniam  [https://www.youtube.com/watch?v=gYhyukRhk1Q] * ZK10: Analysis of zkVM Designs - Wei Dai & Terry Chung  [https://www.youtube.com/watch?v=tWJZX-WmbeY] * Ligerito: A Small and Concretely Fast Polynomial Commitment Scheme [https://baincapitalcrypto.com/ligerito-a-small-and-concretely-fast-polynomial-commitment-scheme/] * Ligero++ - Reducing proof length of Ligero [https://dl.acm.org/doi/10.1145/3372297.3417893] * Adding Zero-Knowledge to STARKs - Talk by Ulrich Haböck [https://www.youtube.com/watch?v=D5nKncWi_Hk&list=PLOEty2U8Y69V6_OJNHTFiQPKH9zQlsBZa&index=23] * Aurora - comparing prover times of STARKs vs Ligero [https://eprint.iacr.org/2018/828.pdf] * WYSTERIA: A Programming Language for Generic, Mixed-Mode Multiparty Computations [https://www.ieee-security.org/TC/SP2014/papers/Wysteria_c_AProgrammingLanguageforGeneric,Mixed-ModeMultipartyComputations.pdf] * Samaritan: Linear-time Prover SNARK from New Multilinear Polynomial Commitments [https://eprint.iacr.org/2025/419] * Brakedown: Linear-time and field-agnostic SNARKs for R1CS [https://eprint.iacr.org/2021/1043] * Intro to MPC-in-the-Head [https://csrc.nist.gov/csrc/media/Projects/post-quantum-cryptography/documents/pqc-seminars/presentations/13-intro-mpc-in-the-head-05212024.pdf]   ----------------------------------------   Check out the latest jobs in ZK at the ZK Podcast Jobs Board [https://jobsboard.zeroknowledge.fm/].    ----------------------------------------   **If you like what we do:** * Find all our links here! @ZeroKnowledge | Linktree [https://linktr.ee/zeroknowledge] * Subscribe to our podcast newsletter [https://zeroknowledge.substack.com] * Follow us on Twitter @zeroknowledgefm [https://twitter.com/zeroknowledgefm] * Join us on Telegram [https://t.me/+taXrPJIp_30zZTNi] * Catch us on YouTube [http://www.youtube.com/channel/UCYWsYz5cKw4wZ9Mpe4kuM_g]   **Support the show:** * Patreon [https://www.patreon.com/zeroknowledge] * ETH - Donation address [https://etherscan.io/address/0xE2C080047213C1d8cDf2099E0B07479C5D9cee8a] * BTC - Donation address [https://www.blockchain.com/explorer/addresses/btc/bc1q9mrh34n6mvses59r3hq9dz6j3vxm3tlwlg8ws2] * SOL - Donation address [https://explorer.solana.com/address/Eqfm4maSDUN3ikjDfSgDeJ1mh3iC8tr69eHB5iTHz35x] Read transcript [https://github.com/ZeroKnowledgefm/episode-transcripts/blob/main/episodes/Episode%20369.txt]

In this episode, Anna Rose [https://x.com/AnnaRRose] and Nico Mohnblatt [https://x.com/nico_mnbl] speak with Dan and Sinu [https://x.com/sinu_eth] from TLSNotary [https://tlsnotary.org/] to trace the project’s journey from early Bitcoin forum ideas to its current role as a foundational protocol maintained by PSE [https://pse.dev/]. Dan recounts the origins of TLSNotary as a tool for cryptographically proving web data, while Sinu explains how the project was revived to provide modern TLS attestation. The conversation covers the use-cases for verifiable web proofs, the different modes these interactive protocols can take, and the broader impact of this technology on leveraging siloed user data. Related links: * Episode 325: Web Proofs with Tracy from Pluto [https://zeroknowledge.fm/podcast/325/] * Episode 354: The Founding of Zero Knowledge Systems with Austin Hill [https://zeroknowledge.fm/podcast/354/] * Episode 362: zkTLS with Maddy from Reclaim [https://zeroknowledge.fm/podcast/362/] * 2013 Blog Post ‘tlsnotary - cryptographic proof of fiat transfer for p2p exchanges’ [https://bitcointalk.org/index.php?topic=173220.0] * TLSNotary Whitepaper [https://tlsnotary.org/TLSNotary.pdf] * DECO: Liberating Web Data Using Decentralized Oracles for TLS [https://arxiv.org/abs/1909.00938] * Primus Labs (Previously PADO) [https://primuslabs.xyz/] * Town Crier: An Authenticated Data Feed for Smart Contracts [https://eprint.iacr.org/2016/168]   ----------------------------------------   Check out the latest jobs in ZK at the ZK Podcast Jobs Board [https://jobsboard.zeroknowledge.fm/].    ----------------------------------------   **If you like what we do:** * Find all our links here! @ZeroKnowledge | Linktree [https://linktr.ee/zeroknowledge] * Subscribe to our podcast newsletter [https://zeroknowledge.substack.com] * Follow us on Twitter @zeroknowledgefm [https://twitter.com/zeroknowledgefm] * Join us on Telegram [https://t.me/+taXrPJIp_30zZTNi] * Catch us on YouTube [http://www.youtube.com/channel/UCYWsYz5cKw4wZ9Mpe4kuM_g] **Support the show:** * Patreon [https://www.patreon.com/zeroknowledge] * ETH - Donation address [https://etherscan.io/address/0xE2C080047213C1d8cDf2099E0B07479C5D9cee8a] * BTC - Donation address [https://www.blockchain.com/explorer/addresses/btc/bc1q9mrh34n6mvses59r3hq9dz6j3vxm3tlwlg8ws2] * SOL - Donation address [https://explorer.solana.com/address/Eqfm4maSDUN3ikjDfSgDeJ1mh3iC8tr69eHB5iTHz35x]   Read transcript [https://github.com/ZeroKnowledgefm/episode-transcripts/blob/main/episodes/Episode%20368.txt]

In this episode, Anna Rose [https://x.com/AnnaRRose] and Nico Mohnblatt [https://x.com/nico_mnbl] speak with Goblin Oats [https://x.com/goblinoats] from Tonk [https://tonk.xyz/] and grjte [https://x.com/grjte] from Bain Capital Crypto [https://baincapitalcrypto.com/] to explore the emerging world of local-first architecture. Goblin shares the evolution of Tonk from an experimental lab working on blockchain games to a product that enables anyone to build personalised local-first software using natural language. grjte discusses her Groundmist project exploring personal data lakes and the intersection between local-first systems and the AT Protocol. The conversation explores the technical foundations of local-first software, including Conflict-free Replicated Data Types (CRDTs), parallels with blockchain systems, and how ZK proofs show up within cloud-optional architecture. Related links: * Tonk GitHub [https://github.com/tonk-labs/tonk] * Groundmist [https://groundmist.xyz/] * Dappicom — NES Emulation in Noir [https://github.com/tonk-labs/dappicom] * Speakeasy [https://tonk.substack.com/p/speakeasy] * Automerge CRDT [https://automerge.org/] * AT Protocol [https://atproto.com/] * Ink & Switch: Local-first software [https://www.inkandswitch.com/essay/local-first/] * Homomorphically Encrypting CRDTs [https://jakelazaroff.com/words/homomorphically-encrypted-crdts/] * Beelay - A new sync protocol for Automerge [https://github.com/automerge/beelay/blob/main/docs/protocol.md] * Jess Martin - The last mile of Local First [https://www.youtube.com/watch?v=VCT4J4g2Ung] * Local-First Conf 2025 [https://www.localfirstconf.com/]   ----------------------------------------   Check out the latest jobs in ZK at the ZK Podcast Jobs Board [https://jobsboard.zeroknowledge.fm/].    ----------------------------------------   **If you like what we do:** * Find all our links here! @ZeroKnowledge | Linktree [https://linktr.ee/zeroknowledge] * Subscribe to our podcast newsletter [https://zeroknowledge.substack.com] * Follow us on Twitter @zeroknowledgefm [https://twitter.com/zeroknowledgefm] * Join us on Telegram [https://t.me/+taXrPJIp_30zZTNi] * Catch us on YouTube [http://www.youtube.com/channel/UCYWsYz5cKw4wZ9Mpe4kuM_g] **Support the show:** * Patreon [https://www.patreon.com/zeroknowledge] * ETH - Donation address [https://etherscan.io/address/0xE2C080047213C1d8cDf2099E0B07479C5D9cee8a] * BTC - Donation address [https://www.blockchain.com/explorer/addresses/btc/bc1q9mrh34n6mvses59r3hq9dz6j3vxm3tlwlg8ws2] * SOL - Donation address [https://explorer.solana.com/address/Eqfm4maSDUN3ikjDfSgDeJ1mh3iC8tr69eHB5iTHz35x] Read transcript [https://github.com/ZeroKnowledgefm/episode-transcripts/blob/main/episodes/Episode%20367.txt]