Timechain
Search Donate
Sponsors
Cake Wallet, private Bitcoin and Monero walletSideShift.ai, no-signup crypto exchangeOrangeRock, non-custodial bitcoin spot and perps tradingBraiins Hashpower, buy and sell Bitcoin hashrateLayer Two Labs, drivechains for Bitcoin
S17 · E27 · 1:42:41

S17 E27: Kieran Mesquka on Railgun & Building Privacy Where the Users Are

29 May 2026 · by Vlad Costea · 12 min read
Published at block 951,550
MP3 (77 MB)

The audio keeps playing while you browse the site.

Read the transcript12,058 words, every line links to the minute it was said

Ask this episode

Embed this episode

Paste this on your site or blog. It is a small player that streams from here: no trackers, no cookies, no ads.

Preview the player

Sometimes you have to wait a year for an interview to happen. I first met Kieran Mesquka at DevConnect in Argentina, back when I was trying to interview both the Tornado Cash people (who still don’t want to make public statements) and the Railgun contributors. Kieran said he was busy, I found other people to interview in the meantime, and life went on. But good conversations are worth waiting for, and this one finally happened: a deep dive into one of the most interesting privacy projects in crypto, built not on its own chain, but right where the users already are.

Kieran is a contributor to privacy projects and best known for his work on Railgun, the zero-knowledge privacy system for Ethereum. But his story starts much earlier, in the most cypherpunk way possible: as a high school kid in 2011-2012 writing and modifying GPU mining clients for Litecoin, squeezing hashes per second out of what he describes as “very unserious hardware,” including a thermally-throttled old MacBook Air. Intellectually stimulating, to say the least.

His Bitcoin discovery story is equally organic: hanging around online communities adjacent to cypherpunk spaces, he stumbled upon casual mentions of this “magic internet money” thing. His first instinct was adversarial: if there’s no central server, can I just find the database file on my computer and add a couple of zeros to my balance? The answer, of course, was no, and understanding why became the rabbit hole: distributed consensus through hashing. That curiosity led to writing what he calls “very electricity wasteful code,” mining clients that worked, generated hashes, did the thing.

The transition to Ethereum came through the same pattern that seems to define his career: toy projects that become serious later. He watched the original ICO of ICOs with skepticism (at the time, BitcoinTalk was pumping out new premine coins with tweaked parameters multiple times a day, so why would this be different?), missed what made Ethereum special at first, then downloaded Parity after launch and found Remix embedded directly in the node interface. No need to fork an entire C++ codebase like on Bitcoin or build metadata hacks like Counterparty. Just script. Cool toy first, serious work later.

From Transparent Money to Private Money

What pushed Kieran toward privacy work is something every crypto user has experienced but few think about seriously. You pick up the bar tab, a friend says “send me some crypto to this address,” and the first thing everyone does is open a block explorer and snoop. It’s cute when it reveals degen memecoin trading. It stops being cute when it’s your daily life: “Vlad, I assume you don’t particularly like broadcasting to the world that you live in whatever neighborhood and you go buy coffee at this particular coffee shop at 8 a.m. every day.”

On UTXO chains like Bitcoin, wallets mostly use addresses once, so piecing together someone’s activity takes some digging. On Ethereum, the account model makes it trivial: one address, everything right there. Add ENS domains on top, and people literally wear their entire financial history in their Twitter handle. Sure, some do it deliberately for transparency (Vitalik being the famous example, though Kieran doubts he books his flights and hotels from that public address), but embedded in that choice is the assumption that a choice exists at all.

Once Kieran started understanding what this technology was actually for, beyond rage-filled forum threads about whose fork of a fork with a 5% different block time is best, the question became obvious: why would anyone buy coffee, or take out a future home loan, on a system where everyone they ever interact with can look up all the details? If you want this stuff to be generally accepted, sensitive financial operations need to happen privately.

Railgun vs. Tornado Cash: Privacy Where the Usage Is

So what’s the difference between Railgun and Tornado Cash? Both use ZK-SNARKs, both aim to bring Zcash-grade privacy to Ethereum. Kieran’s answer is that Railgun, which has been around for about six years now, was built around a specific insight: the most valuable place for privacy to exist is right alongside where the usage you want privacy for already exists. Earlier options forced a choice: either you got a very limited use case (deposit, hold, withdraw somewhere else later), or you had to leave for a completely different chain. Asking people to cut themselves off from Ethereum’s rich DeFi infrastructure just to hodl privately is a pretty big ask.

Functionally, Railgun creates shielded pools comparable to Zcash’s Orchard, adapted to the constraints of the EVM. But the key differentiator is expressiveness: Railgun contracts can bind arbitrary additional data into transactions, enabling multicalls, sequences of actions executed from a private balance. Mechanically: you have UTXOs in the shielded pool, you sign a transaction and generate a ZK proof, pop your 100 DAI into a multicall address, route it through your favorite DEX, receive ETH, and shield it back as a new UTXO.

Swaps on Uniswap, supplying LP, whatever DeFi thing you want to do, all from a private Railgun balance. It’s a full wallet transaction system, meant to be the set of protocols a regular wallet can use, not some separated island off somewhere else.

Why the Tornado Cash Devs Got in Trouble (and Railgun Didn’t)

The unavoidable question: why did Alexey Pertsev and Roman Storm end up in legal trouble while Railgun operates freely? Kieran, not a lawyer, wisely declined to speculate much, offering only the timeless advice: “don’t be stupid online.” I offered my own theory, which mirrors the Zcash vs. Monero dynamic. Tornado Cash’s branding leaned into the washing-machine-meme aesthetic, much like Monero’s community embraces darknet market credentials.

Zcash, with arguably better cryptography, chose the “privacy for grandmas, protection from Chinese blockchain analysis, human rights” narrative instead. In Bitcoin, we saw the same split: Wasabi Wallet added a blacklist in cooperation with chain analysis and its devs still roam free, while Samourai Wallet actively tweeted at sanctioned Russian oligarchs to come mix with them, and its devs went to jail. The technology matters less to prosecutors than the speech around it.

That comparison leads to Railgun’s most controversial design choice: PPI (Private Proofs of Innocence), a filtering system that ensures your funds aren’t in a privacy set with known bad actors. A Monero-minded viewer in the chat called it “guilty until proven innocent, sounds backwards and effed up.” Kieran, ever the diplomat, declined the philosophical debate, noting only that the chain isn’t a court of law, and that people hold very strong opinions on this in both directions.

My take: the market provides options for everyone. Wasabi-style filtering gave ordinary users the confidence that they wouldn’t get tainted by sitting in a pool next to a sanctions evader, and Railgun occupies a similar niche on Ethereum: privacy for normal everyday people who want to use DeFi without leaving a permanent public trace of everything they’ve ever done. Zcash, by contrast, lets anyone shield regardless of history. It’s a political difference in the approach to privacy, not a cryptographic one.

Trusting the Moon Math

Speaking of Monero conservatism, I asked Kieran the question that always comes up: why should people trust zero-knowledge “moon math” that almost nobody can personally verify? His answer was the best technical segment of the episode. Historically, privacy protocols faced a fundamental trade-off: if the soundness breaks, either the privacy breaks (Monero’s choice: better deanonymized than inflated) or the validity breaks (the Zcash-family risk: potential secret inflation). That was a reasonable stance for Monero to take at the time.

But the technology has evolved: the constructions are far more battle-tested now, and the trade-off space has widened. Railgun uses Groth16, which remains extremely hard to beat on verifier efficiency: on Ethereum that translates directly into lower gas costs, and on any chain into less state growth and cheaper verification. The sweet spot, in Kieran’s framing, is cryptography that is neither brand-new moon math nor needlessly expensive: reasonably performant, reasonably succinct, reasonably scalable. The days when “ZK is untested” was a strong argument are behind us.

Smart TVs, Robot Vacuums and the War on Convenience

A long and unexpectedly delightful stretch of the episode had nothing to do with crypto at all. Bonding over Naomi Brockwell’s work (the Australian privacy educator who connected us, and who Kieran praises for making privacy accessible through small intentional actions: not handing over your phone number at checkout on autopilot, unchecking the “optional” data collection boxes), we went deep on the surveillance creep of everyday objects. Cars with embedded SIM cards (which some hackers famously extracted for free mobile data), smart TVs running content-ID fingerprinting that screenshots whatever you watch even over HDMI (unplug yours from the internet, or block it in your router), washing machines phoning home, robot vacuums mapping your house, with two glorious horror stories: one vacuum model using an unauthenticated cloud MQTT server, letting strangers drive other people’s vacuums around; another wave where hackers remotely took over vacuums to follow people around their homes shouting slurs through the speakers.

Kieran’s analysis of why this happens is sharper than the usual privacy-absolutist take. These data collection avenues exist because they meet real demand: convenience is a compelling selling point, economies of scale make surveillance the default, and opting out becomes nearly impossible (try buying a dumb TV, a dumb washing machine, or a decent vacuum without an app). The “reject modernity, return to monke” mindset appeals to the hardcore few but is a losing battle.

VHS walls are cool; Netflix menus are convenient. If you want to build a better world where people get their agency back, you have to build privacy where the usage is going, not convince everyone to wash their clothes by hand. Also: printer drivers abandoned by manufacturers, HP’s ink subscriptions that require cloud-connected printers, Logitech’s subscription ambitions for keyboards and mice. Hardware perhaps just needs to cost more again: take my money once and leave me alone.

Why Bitcoin Doesn’t Have This

As a hardcore Bitcoiner hosting what is still a Bitcoin podcast, I had to ask: why did Ethereum get state-of-the-art privacy tech while Bitcoin didn’t? Kieran sees both cultural and technical layers. Culturally, Bitcoin’s prevailing narrative became “accumulate and wait”: if you’re not very active on-chain, you’re not posting much doxxing information, so privacy feels less urgent, and you rationally prioritize stability over feature set. Technically, Bitcoin Script is genuinely limited in expressiveness; you can argue more is possible with it than people assume, but anything too fancy gets very difficult very quickly.

The result: innovation gets pushed to the edges, to islands like Citrea (which, fun fact, hosts a Tornado Cash fork called Crest that almost nobody uses), rather than the base layer. And his quiet killer point: if the resistance were purely technical, Zerocoin would have been implemented a decade ago. The original paper was literally presented as a Bitcoin upgrade. Performance concerns were valid in 2013; after a decade of cryptographic improvement, much less so. The demand just isn’t there culturally, at least not yet. I still hope shielded CSV and the BitVM ecosystem change that.

Opt-In Privacy, Ethereum’s Health, and the Elevator Pitch

To the Monero-maximalist argument that opt-in privacy outs you as suspicious, Kieran offers a pragmatic defense of optionality: the transparency of DeFi is itself valuable. You can audit that a lending protocol is really 2x overcollateralized; that visibility is what separates Ethereum DeFi from someone running the same thing on a central server. Some things belong in the open; some things you don’t want bare on the internet. Having the option for both is the point.

Is Ethereum dying, as this week’s grave-dancers (a Bankless host selling his ETH, EF researchers departing) claim? Kieran checked Etherscan live on stream: no meaningful drop-off in activity. Public turmoil around the Foundation won’t kill the chain in weeks or months, if ever. “Ethereum is the DeFi king of chains. It has a fair ways to go before it is dethroned, if it is dethroned.” Railgun, meanwhile, already lives on Arbitrum, BSC and Polygon (with modest usage outside mainnet), and future deployments follow usage. Solana has some interesting privacy features in its token standard, but building something as complicated as Railgun on the Solana VM is hard, and productive DeFi demand there remains an open question.

And the one-minute elevator pitch for a conference stranger? “Railgun is a protocol that allows you to have wallets that can do whatever you can do with a public wallet, just privately.” Shorter than I expected, but then again, the best tools are simple to describe and thorny under the hood.

The Future: PIR, Smart Wallets and Better Entry Points

For the closing question, courtesy of the chat: what privacy features don’t exist yet but live in Kieran’s head? Two answers. First, private information retrieval (PIR) standards: retrieving your balance privately is computationally heavy compared to a single query on a public chain, and solving that UX bottleneck would lift the entire privacy space. Second, new Ethereum EIPs like 7702 and frame transactions, which make the entry point of transactions programmable: historically, validity meant a correct signature plus enough ETH for gas; smart-wallet-style programmability changes that, and Railgun effectively already acts like a smart wallet in this paradigm, making privacy usability a whole lot nicer.

Also addressed along the way: the “Kieran is Satoshi” running joke (he finds it mildly annoying, especially when people at meetups don’t realize it’s a joke; he is fortunately not the first Australian to be called Satoshi), the giveaway of goodwill from a chat full of Railgun fans, and the fact that he doesn’t watch price charts because life is more zen that way.

Kieran doesn’t do much social media, but you can find him on X/Twitter and GitHub as @mesquka. Conference appearances are unplanned but probable; any talks will surface on Twitter. Follow his work, try a privacy wallet, and remember: the biggest privacy gains start with small intentional actions, like not giving your phone number to the checkout clerk.

This episode is sponsored by Orange Rock (trade without KYC or registration, leverage optional and not recommended), Sideshift.ai (swap your stablecoins for free market money before they get frozen, as millions in USDT and USDC were just last month), Cake Wallet (the privacy wallet with Monero, Zcash, silent payments, PayJoin and swaps), and LayerTwo Labs (drivechains, the Zside chain, and the Ecash fork launching this August at block 964000: check ecash.com, every Bitcoin self-custodian gets free coins). Special thanks to all the sponsors for making this episode freely available to everyone!

Listen to Kieran Mesquka on YouTube, Spotify, Apple Podcasts & more!

Watch on YouTube

Time stamps:

00:01:10 – Intro: Kieran Mesquka, one year in the making
00:02:21 – Writing GPU mining clients for Litecoin in high school
00:03:54 – Discovering Bitcoin: can I just edit the database file?
00:06:31 – From mining to Ethereum: the Parity node with built-in Remix
00:10:24 – Why privacy? The bar tab problem & your coffee routine
00:16:18 – ENS domains: your entire financial history in a Twitter handle
00:20:33 – Railgun vs Tornado Cash: privacy where the usage is
00:24:38 – Multicalls & DeFi from a shielded balance
00:26:30 – Why did the Tornado Cash devs get in trouble?
00:29:10 – The “Kieran is Satoshi” joke & other Australians
00:31:07 – Naomi Brockwell & privacy as small intentional actions
00:33:25 – Surveillance cars, kill switches & embedded SIM hacks
00:35:47 – Smart TVs screenshotting what you watch
00:39:06 – Smart washing machines, vacuums & printer driver rage
00:41:17 – Robot vacuums: MQTT hijacks & slur-shouting hackers
00:44:01 – Subscriptions for hardware: take my money once, leave me alone
00:46:36 – Build privacy where the usage is, not “return to monke”
00:51:04 – Sponsors: Orange Rock, Sideshift, Cake Wallet, LayerTwo Labs
00:54:43 – Is Railgun’s privacy equal to Zcash’s? The Orchard comparison
00:57:15 – How a private DeFi transaction actually works
01:00:00 – Opt-in vs enforced privacy: the Monero maximalist argument
01:02:38 – Why doesn’t Bitcoin have this tech? Culture & expressiveness
01:08:26 – The Zerocoin point: it was proposed as a Bitcoin upgrade
01:10:33 – Eleanor’s question: Railgun on other chains (Arbitrum, BSC, Polygon)
01:12:42 – What about Solana? Token standard privacy & dev difficulty
01:15:05 – Is Ethereum dead? Checking Etherscan live on stream
01:19:13 – The one-minute Railgun elevator pitch
01:20:48 – PPI filtering: “guilty until proven innocent” criticism
01:24:28 – Wasabi vs Samourai: the Bitcoin parallel for filtering
01:27:37 – Privacy for everyday people vs privacy for criminals
01:29:20 – Trusting the moon math: why ZK skepticism is outdated
01:31:40 – The soundness trade-off: privacy breaks vs inflation breaks
01:33:57 – Groth16 & verifier efficiency: the scalability sweet spot
01:36:18 – Privacy tech that doesn’t exist yet: PIR standards
01:38:09 – EIP-7702, frame transactions & programmable entry points
01:40:36 – Closing: why Ethereum is more open to base-layer privacy
01:41:03 – Where to find Kieran: X & GitHub @mesquka
01:42:36 – Outro & sponsor thanks

Verify this episode

This is the SHA-256 fingerprint of the MP3 as published here, timestamped in the Bitcoin blockchain with OpenTimestamps. If one byte of the file changes, the fingerprint changes.

SHA-256
f17f51b958f7de8cdb210569a9b0f1d77c436316500be5527f64fec6aad2b6ab
Proof
2026-06-01_s17_e27_kieran_mesquka_on_railgun_building_privacy_final.mp3.ots · all episode hashes (proof)
Status
Anchored in Bitcoin block 974,312
How to check it yourself
  1. Download the MP3 and the proof into the same folder.
  2. Run sha256sum 2026-06-01_s17_e27_kieran_mesquka_on_railgun_building_privacy_final.mp3 (on a Mac: shasum -a 256 2026-06-01_s17_e27_kieran_mesquka_on_railgun_building_privacy_final.mp3). The result must match the SHA-256 above.
  3. Install the OpenTimestamps client (pip install opentimestamps-client) and run ots verify 2026-06-01_s17_e27_kieran_mesquka_on_railgun_building_privacy_final.mp3.ots. It hashes the MP3 again and checks the proof against your own Bitcoin node, then prints the block and the date the file existed by. Without a node, ots info 2026-06-01_s17_e27_kieran_mesquka_on_railgun_building_privacy_final.mp3.ots shows the block height to look up in any block explorer.

Share: X · Email · Link

Related

Comments

Comments run on my own server. No ads, no trackers.

0:000:00