S15 E37: Marek Narozniak on MimbleWimble and Grin

Some guests bring a sales pitch. Marek Narozniak brought a PhD thesis on quantum computing, a half-finished Python implementation of a privacy protocol that most of crypto Twitter buried in 2020, and the calm patience of a physicist who has heard every "quantum will kill Bitcoin" headline and stopped flinching.

Marek is a Polish computer scientist who spends his after-hours on Grin, the first implementation of the MimbleWimble protocol. His day job is Bitcoin-adjacent too: he runs a service focused on transaction monitoring, lookups and custody tooling. So when I invited him to my show, I had two reasons. The first was the proposal to bring MimbleWimble to Bitcoin through an extension block, the same trick Litecoin pulled off with MWEB (if that sounds like Greek, start with my previous conversation with Charlie Lee on MWEB). The second was the eternal quantum boogeyman.

As I told Marek at the start, "we're not talking about something that people already know." This was an episode about unorthodox expertise, and he delivered.

The Privacy Model That Was Never Broken

I opened with the elephant in the room. In early 2020, a report claimed Grin's privacy had been broken, and attention drifted toward shinier things like Halo 2 on Zcash. Marek corrected me right away.

"Actually the privacy model was not broken," he said. What the research showed is transaction linkability when the network is quiet: if a block carries a single transaction, an observer can connect the dots even though there are no on-chain addresses and no visible amounts. And that, in his words, "remains true for literally any cryptocurrency no matter how private the protocol is."

This is the anonymity set problem. Privacy is not just cryptography, it is hiding in a crowd. A small crowd makes you easy to spot. A big one makes you a ghost.

Harry Potter, Pedersen Commitments & Cut-Through

So what is MimbleWimble? Marek described it as "a blockchain protocol for peer-to-peer cash" that hides amounts, with transactions stored cryptographically in Pedersen commitments. Grin is simply the first implementation of it.

He credited the design to bitcoiners around 2015, Andrew Poelstra included, and I couldn't resist adding the lore. The original idea came from a Satoshi-like figure using the pseudonym Tom Elvis Jedusor, the French name for Lord Voldemort. Before that, the main privacy proposal in Bitcoin land was Confidential Transactions, which bloated transaction sizes and fed the belief that privacy doesn't scale. The MimbleWimble paper flipped that script. Marek was gracious about it: "you're actually better prepared than me."

The magic trick is cut-through. Transactions get aggregated, and what remains per transaction is mostly the kernel, kept in an MMR data structure (he had double-checked this with Grin co-founder John Tromp that very day). I had a slightly different mental model of how much gets pruned, and Marek politely told me he wasn't sure about my version. Fair enough. The result, by his numbers, is a chain whose growth rate is "four times smaller than Bitcoin and 20 times smaller than Monero." When I played the critic and suggested that's just because nobody uses Grin, he pointed out the comparison tables assume equal transaction loads.

A Lockdown Hobby & a 90-Minute Full Node

Why would a quantum computing PhD pick Grin out of every project on earth? Lockdowns. Marek was ahead on his papers, bored, and a fellow PhD student told him about "this like super cool protocol." He got pulled in, and it stuck as "my like after hours hobby kind of thing." He is now building his own Python implementation, which can already construct a valid transaction.

I shared my own experiment: I synced a full Grin++ node in 90 minutes. Marek said that is "not bad given the blockchain is 5 years old," but noted Grin++ is an older codebase that lacks newer tricks like PIBD (parallel initial block download). He also reminded me of the original point of running a node: being your own bank. Back when David Burkett released Grin++, "it was the first cryptocurrency that could actually do it" on a phone. David, of course, went on to build Litecoin's MWEB, and Marek still pings him when his own code misbehaves. (You can hear David himself in S15 E59 on MWEB & Privacy.)

99 Bitcoin in the Bank, Nobody to Invoice

There is a popular story that Grin died because it can't pay developers. Marek begged to differ. Funding requests on the Grin forum are "very well paid actually," and the community budget sits at something like 99 bitcoin.

The real problem is boring paperwork. Grin is a community project with no official organization to invoice, which turns accounting into a nightmare. That is exactly why Marek has never applied: "I prefer to do some work for free rather than go through you know hell of trying to declare income from you know some unregistered source." His open invitation stands: if you know Rust and like MimbleWimble, "there is money and there is you know work to do."

CoinJoins, MWixnet & the Monero Comparison

I asked about CoinJoins on Bitcoin. Marek was refreshingly honest: he hasn't tried them and gets his coins in ways that don't require them. His mental image comes from memes where Grin goes straight from one person to another while Bitcoin needs "some other acrobatics."

On Grin, aggregation is native, so you can technically combine transactions yourself. The hard part is doing it non-interactively through a third party, since Grin builds transactions interactively with slatepacks, send flows and invoice flows. Enter MWixnet, now in phase three of implementation by an anonymous developer. It lets users publish partial transactions and collectively pay to have them merged. Once it is live, Marek believes Grin's privacy would be "equivalent to Monero as much as I understand it," still with that tiny chain growth.

Grin vs Bitcoin Philosophy

I reminisced about the 2019 launch hype: Bitcointalk accepting Grin donations, Jameson Lopp tweeting about running it, someone swapping one Grin for one bitcoin on launch day. Then the price faded. So why should bitcoiners care?

Marek blamed part of it on liquidity and delistings, and part of it on Grin's deliberate design. Emission is linear and forever: "one coin per second." That meant brutal early inflation, now below 20% and falling. Grin, as he put it, is "not meant to transfer money from like new people into the founders." His verdict depends on who's asking. If you came to Bitcoin for decentralized cash and fairness, you'll find a lot to love. If you want a rocket, look elsewhere.

He also gave a fair-minded take on Beam, the commercial sibling: a company structure, non-interactive transactions, a DeFi and NFT appetite, a premine and no linear emission. NFTs on a privacy chain sounded oxymoronic to me, but Marek explained you can create MimbleWimble tokens by using different curve generator points, so "if you can make tokens you can also make NFTs in a way."

Can Bitcoin Get a MimbleWimble Extension Block?

Technically, Marek sees no obstacle: "if it's possible on Litecoin I don't see the reason why it wouldn't be possible on Bitcoin." Politically is another story. I argued Litecoin has a creator who still steers the ship, while Bitcoin is a leaderless bazaar where developers currently want covenants and users don't understand why they'd want opcodes they'll never touch.

I floated putting MimbleWimble on a drivechain or other trust-minimized sidechain (a topic Paul Sztorc refused to name in his own episode). Marek's one condition: same proof of work, same consensus, no sneaking in other agendas. "I'm not advocating for any changes in that."

Quantum FUD, Dismantled Calmly

Then the main event. Marek sees two attack vectors: quantum mining and breaking signatures. The scary one is using Shor's algorithm (a modified version works on elliptic curves) to derive private keys from public keys.

His good news: Bitcoin addresses are hashes, and "quantum computer cannot really solve a hashing function significantly faster than regular computer." The danger window opens when you spend and your public key lands in the mempool. An attacker would need a machine big enough to crack it before the transaction confirms, "really really really big." Before that happens, Bitcoin would have time to adopt quantum-safe cryptography. The genuinely exposed coins are those with revealed public keys, including some famous addresses, plus anyone who shares an extended public key.

On qubit headlines, he was blunt. A thousand-plus physical qubits translates to something like nine logical qubits of reliable computation. I added my cynical take that big quantum announcements are fundraising tools: if they could crack Bitcoin, why not grab Satoshi's coins? Marek's own warning was that quantum computers are "not solution to everything." They excel only at problems tied to the Schrödinger equation.

His unpopular opinion was my favorite nugget. Classical computing didn't win by building ever-bigger circuits, it won with the von Neumann architecture: CPU plus memory. Quantum could do the same. With reliable quantum memory, "you could make a reliable quantum computer with just two reliable qubits," paying only a quadratic penalty. "You don't actually need more qubits you need quantum memory."

When I pushed for a clickbait number, he guessed "maybe 20 years," not this decade, and in his lifetime (he's 34). By then, he believes quantum-safe schemes like Lamport signatures will bring "fresh ideas and applications as well." For a deeper dive into the post-quantum debate, see my later chat with Hunter Beast & Mark Karpeles.

Grin Nodes for Kevin

A viewer named Kevin chimed in from the chat: he mines Grin but his Ironbelly wallet's node was down. Marek recommended grinnode.live, a public node he helps manage, started back in 2020 when running nodes was a struggle. Still, "you should run your own node." I told Kevin the same, with feeling: it takes 90 minutes. We also had a small detour into Raspberry Pi nodes, where I confessed my Pis have all crashed at some point and a used laptop is often more reliable.

Bitcoin in Poland: Hoodies, Invoices & Bankers Pouring Coffee

Marek holds his savings in Bitcoin, calling it worth owning "especially in uncertain times," with fear of war coming from the East. He doesn't mean price, he means escape velocity, citing WikiLeaks as the classic case where Visa and wires stopped but Bitcoin kept working. He also invoices clients in bitcoin in Poland, "perfectly legal no issues with that," and wears his Bitcoin hoodie every time he visits a bank. The reactions, he says, "are very positive."

I told him about attending the 10-year anniversary of the Polish Bitcoin Association, hosted at a bank's headquarters, where "the bankers were the ones serving them coffee." Marek admitted he keeps postponing the Warsaw meetups and promised I'd motivated him to go.

A Government Signature Service, Accidentally Cypherpunk?

The finale was a gentle sparring match. Poland offers a government service for signing any XML file with your ID, and Marek wondered when someone will turn it into an oracle or even a seedless Bitcoin wallet. I countered that signing under your real identity sounds like "a gateway to CBDCs." He proposed zero-knowledge proofs to reveal only what's needed, like proving you're an adult. I called him an optimist. He agreed governments like privacy in the CBDC sense: "privacy between us not privacy like from them." Still, he believes "sometimes government does something cypherpunk by accident" and plans to explore it at the upcoming ZK Hack in Kraków.

My closing pitch was half personal agenda: Bitcoin needs privacy so it doesn't turn into something Orwellian. Grin is not dead. It is a privacy project with the promise of scalability, and that's still a big deal. And as Marek proved, sometimes the most useful research is the one nobody is hyping. You can follow his work at mareknarozniak.com.

Listen to Marek Narozniak on Spotify, Apple Podcasts, YouTube, and More

Watch on YouTube

Time stamps:

00:00:40 - Introduction to MimbleWimble and Grin
00:02:07 - The Privacy Model of MimbleWimble
00:03:47 - What is MimbleWimble and Grin?
00:05:18 - Scalability and Growth of Grin
00:07:16 - Marek's Interest in Grin and Quantum Computing
00:13:36 - Challenges and Funding of Grin Development
00:15:41 - CoinJoin on Bitcoin
00:16:49 - CoinJoin and Privacy in MimbleWimble
00:22:18 - The Relevance of MimbleWimble and Grin to Bitcoin
00:24:15 - Bitcoin Philosophy
00:25:37 - Scalability and Design of Grin
00:27:31 - Compatibility with Bitcoin
00:38:31 - Quantum Computing and Bitcoin
00:45:02 - Quantum Resistance in Blockchain Networks
00:46:27 - Node Software and Wallet Connectivity
00:47:49 - Running a Node and Raspberry Pi
00:50:02 - Quantum Computing and Qubits
00:52:07 - Quantum Computers and Bitcoin
00:52:42 - Bitcoin and Cryptography
00:56:01 - Bitcoin and Investment
00:57:31 - Bitcoin Adoption and Use Cases
01:02:46 - Polish Bitcoin Association and Government Perception
01:04:32 - Government Digital Signature and Privacy
01:05:44 - Privacy and Bitcoin