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
S16 · E32 · 5:08:14

S16 E32: Bitcoin vs Monero vs Zcash Privacy with Diego Salazar & Luke Szramowski of Cypher Stack

3 Jul 2025 · by Vlad Costea · 13 min read
Published at block 903,844
Cover art for S16 E32, Bitcoin vs Monero vs Zcash Privacy with Diego Salazar & Luke Szramowski of Cypher Stack
MP3 (423 MB)

The audio keeps playing while you browse the site.

Read the transcript51,532 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

"The reality is that most people don't threat model," Diego Salazar told me about forty minutes into a five-hour livestream, and the rest of the evening was built on that sentence. Diego founded Cypher Stack, builds Stack Wallet, and has spent years working on privacy across more than one coin. Next to him sat Luke Szramowski, a mathematician at Cypher Stack who works on the research behind Monero's FCMP upgrade. I had asked the pair to do something nobody on this show had tried before: sit down with me and rate every major privacy system, one number per property, until we had a table we could argue about.

The rubric started with five criteria that I brought: sender, receiver, network, amounts and metadata. Diego added linkability about fifty minutes in, then footgunability at the one hour fourteen mark, and I tacked on scalability around one twenty-five. We scored Bitcoin (with CoinJoin, Silent Payments and BIP47 thrown in), Monero before and after FCMP, Zcash shielded and transparent, Firo, and Litecoin with MWEB. By the end we had numbers, a mountain of footnotes, and two guests who did not want to be pinned to a single digit.

Somewhere around hour five I praised their iron bladders, and at hour two I had held up a laptop and a stack of burned DVDs from the 2016 Zcash ceremony, which I bought from Peter Todd for $1,000.

Nobody Is Threat Modeling

I opened by asking how they decide what a normal person needs, and Diego's answer was that normal people have not decided anything. They install an app, send some coins and assume the rest is handled. He compared the habit he wants to build to "defensive driving": you assume the other driver is the problem. For privacy that means assuming that the one watching you is patient, well funded and paid to correlate. "It's not me that I need to look out for," he said, and the point he was making is that your own leaks are only half of it.

His priority list was blunt. "Metadata is the thing that kills," he said, meaning IP addresses, timing, the exchange that sees both ends of a withdrawal, the device you used. He has watched people obsess over the cryptography of a transaction and then broadcast it from their home connection.

Luke's version came with an image. Using a privacy tool without covering the network layer, he said, "wouldn't make sense to put on an invisibility cloak if you're just going to hold up a police light above your head outside the cloak, right?" Later in the evening he gave me an example from his own reading, an acoustic side-channel attack that recovered passwords from the small differences in sound between keystrokes, picked up through a microphone. That attack has no interest in your ring signature.

I also owned up to an error early on. Explaining why encryption standards matter, I said AES stood for Advanced Encryption Symmetry, and the chat corrected me within seconds. It is Advanced Encryption Standard. Luke took it well, and told the audience that the only shame is in pretending to know something you do not: "I'm an idiot, but I'm an idiot that's good at a handful of things, and I think that's good enough."

What CoinJoin Actually Buys You

I scored Bitcoin first, with the generous assumption that you do everything right: CoinJoin, Silent Payments for receiving, BIP47 payment codes, the works. Diego started at zero for sender privacy. His reasoning is that Bitcoin's design publishes the spending graph, and anything you add sits on top of it.

He is not dismissive of CoinJoin. "It's better than nothing," he said, and he later added that "there is an emergent property with CoinJoin that makes the privacy stronger than the sum of its parts," because chained rounds compound the doubt for an observer. What he objects to is the marketing. He recalled a conversation at a 2018 conference where the confidence around CoinJoin outran anything he could back up with numbers, and said he has been hearing the same confidence ever since.

Luke's image for the structural problem was a good one: "This sort of CoinJoin almost feels like a suit of armor that you're putting on top of like skin." The armor is real and it helps, but it was bolted on, and you can see the seams. A system designed for privacy from day one has no seams to find.

I pushed back on the receiver score. With Silent Payments, I argued, receiving can be cleaner than the zero we were headed for, and we settled on a range of 0 to 2 for Bitcoin there, depending on how disciplined the user is. Amounts got a 3 and metadata a 1. For network privacy I gave Bitcoin a 3, since you can run Tor and nothing stops you, but nothing helps you either. Non-footgunability went to 0: the ways to ruin yourself by accident are everywhere, from address reuse to merging the wrong coins.

Max Hillebrand, who has spent years on Bitcoin privacy advocacy, was on the show earlier in S6 E6: Max Hillebrand on Bitcoin Privacy Advocacy & Wasabi Wallet. Diego mentioned WabiSabi, the protocol behind the coordinator model, among the more serious attempts to make CoinJoin less awkward.

The Footgun Column

Footgunability was Diego's addition, and it ended up deciding more arguments than any other column. He framed it as a question: "how easy is it to just completely shoot yourself in the foot and just ruin all of your privacy."

His own experience with Bitcoin is a useful example. "Bitcoin used to be so simple," he said, and then he walked through the modern version: a coordinator, a round that has to fill, a wait, and a final step to "press send about 20 minutes later." Every extra step is a chance to do the wrong thing in the wrong order. Luke put the principle plainly: "You can't expect someone to reasonably defend themselves against something that they don't even know exists." A privacy system that depends on the user knowing a list of rules will fail users who do not know the list.

Diego also brought up his PGP key, to show where the pain sits: "I have a PGP key, man. I rarely dust that thing off and use it because it sucks, man. It sucks to use is everything sucks about it." Luke added a historical case, the Enigma break, where German operators signed off every message the same way and thereby shrank the search space for the people trying to read them.

That column is where Monero scored best, a 9, because the defaults do most of the work and there is little for the user to configure. Litecoin with MWEB got a 7. I gave Zcash less certainty here, and we came back to that when we reached the transparent addresses.

Footnotes accumulated fast. A score of 9 for Monero's footgun resistance still sits next to a 3 for its sender privacy, which brings me to the next section.

Monero Before and After FCMP

Diego does not believe in sanding down the origin story. "Monero started off as a scam, too," he said, referring to Bytecoin, the CryptoNote coin that Monero's early developers forked after concluding that the premine was rigged. It was his way of telling me that a coin can begin badly and become the thing everyone else in this space is measured against.

The current sender number is where the discomfort begins. "You're hiding among 16 outputs," he said, and the arithmetic is unforgiving: a ring of 16 is a small crowd, and chain analysis firms have spent years learning how to shrink small crowds. I gave Monero a 3 for sender privacy before FCMP, and Diego did not argue. Receiver privacy got an 8, thanks to stealth addresses, and amounts a 7. Network privacy came in at an 8, which reflects Dandelion++ and the fact that the project took the question seriously. Metadata got a 6, and non-linkability a 6.

Full-chain membership proofs change the sender story. Luke spent a stretch of the show on the divisor research that makes FCMP possible, and he was open about how long the work has taken. "a much greater appreciation for just the sheer time commitment that privacy research requires," he said of the experience. He also described himself as "the himbo of the research group," which I am including here because it was the funniest thing said all night. On timing, Diego said the upgrade is expected "maybe this year or at the very least Q1 next year."

With FCMP, I scored sender at 9 and non-linkability at 8 to 9. The cost is on the scalability side: verification gets heavier, so Monero moves from a 4 to a 3. Luke's comment that "higher on average doesn't necessarily equate to better all the time" is a fair summary of what the rubric can and cannot say.

I later recorded an episode with Riccardo Spagni on Bitcoin, Monero and privacy: S17 E8: Riccardo Spagni on Bitcoin, Monero & Privacy. The Monero project keeps its research notes public.

Zcash, Its Transparent Addresses, and a Ceremony in a Box

For shielded Zcash, the scores were as high as anything on the board: sender 10, receiver 10, amounts 10 and metadata 8 to 9. Diego grants it the cryptography without hesitation, and the work around Halo 2 and Orchard removed the trusted setup that used to be the usual objection.

He then listed what he thinks it cost. "I feel like they missed their chance to do something really big," he said, and his reason was the transparent side. "The T-addresses are just Bitcoin," he told me, and since many exchanges and wallets only touch those, the majority of activity never gets the shielded benefits. He called the arrangement "a compromise of a compromise." I asked him about network-level privacy in Zcash, because the protocol on its own does not hide your IP, and he agreed that it is the weakest point of the otherwise strong score. Josh Swihart and Kris Nuttycombe talked about Zcash, zero-knowledge proofs and cash for the internet in S15 E62: Josh Swihart & Kris Nuttycombe on Zcash, Zero-Knowledge Proofs & Cash for the Internet, and the current protocol documentation lives at z.cash.

This is where I reached under the desk. A few years ago I paid Peter Todd $1,000 for the remains of the 2016 ceremony, which he said he was clearing out of his closet. "I have all the remains, including the DVDs," I told the chat, and held up the laptop and the burned discs. Diego laughed and said he now has a very specific object he wants to buy from me.

Zcash's usability problem is not the math, in his telling. "Bitcoin has horrible UX," he said when we compared the two experiences, and I pointed out that this is half the reason people reach for wallets like Edge. Luke's own take, after reading the Zcash paper end to end, was that the paper "got everything, which is important," and that understanding all of it would take a month of burning it into his brain.

Firo, MWEB and Everything We Left Out

Firo was the next entry, and Diego's description was short: "pretty much defunct" in terms of developer attention compared to the other projects on the board, though he was careful to say that Reuben Yap's team has shipped real work. I had Reuben on in an earlier episode, and the Firo site has the details on Spark.

What Diego did want to talk about was the exchange delisting, which he treated as a lesson for others. "They played by the rules, they played the games, they did the things, they appeased the people and they got delisted anyway," he said. If Zcash is counting on compliance to buy it safety, the Firo story is the counterexample. He called the privacy that survives only with a regulator's permission "very brittle privacy because it does not take much to completely nullify its effects."

Litecoin with MWEB scored 5 to 7 on sender, receiver and amounts, a 7 for non-linkability and a 7 for non-footgunability, a 3 for network privacy and a 6 for scalability. Diego described Lelantus-MW and Beam in passing.

On the network layer I made a smaller observation. Diego went through why the Tor documentation advises against chaining a VPN before Tor: you add a party that sees your real address and your destination timing, without removing the one that Tor was already protecting you from. He asked me to read it out and then said that most people still do it anyway. "instills a false confidence that people aren't collecting passive stuff about you," he said, and I told him I think that is the biggest weakness in the whole stack. He also pointed to the research on Stack Wallet's FROST multisig implementation, which he feels Bitcoiners ignored entirely.

Everyone on stage was a little worn by this point. I had been calling out "iron bladders" for an hour.

One Number, a Mountain of Footnotes

When the table was full, Diego was the first to say it flattens things. "I'm distilling it down to a single number," he said of each cell, and he added that the table needs "a mountain of footnotes." Luke's approach was gentler: he liked the exercise because it forces the argument into the open, and said "That's an awesome analysis that should be done more." The numbers, as Luke noted, were not pulled from nowhere: "we didn't just pull those out of thin air." Diego's closing request on this part was that people stick to the scores: "stand by those numbers."

The last stretch of the evening was about why any of this matters beyond the table. Luke has heard the "nothing to hide" argument often enough to have a counter-reply, which is that people choosing to keep something private are saying "you don't deserve to see it." He thinks "like 99.9% of people are as true neutral as you can be." Diego was less sunny, offering "I think humans are at most inherently selfish." Both ended up in the same place: financial privacy should not depend on assuming that people are good.

I asked Diego whether he wants Bitcoin to be mainstream without base-layer privacy. "I would not like Bitcoin to become mainstream yet," he said. His argument is that a surveillance-friendly Bitcoin will become the default for the next billion people, and changing it later will be nearly impossible. I stated the opposing view as plainly as I could: "pretty much everything that we're going to discuss could have been part of Bitcoin." Diego's reply started with a line I did not expect: "I am also a potential Bitcoin maximalist," he said, which should surprise people who know him from his work on other coins. He does not view the Bitcoin fix as hopeless, and he thinks there is a way to bring this work to Bitcoin without waiting for permission from a dozen different camps. Shielded CSV is one such effort.

One more thing about how Diego talks. He said of Twitter that "it values wit over nuance," and that is why he spent five hours with me instead. After the last score I asked him for a parting line for anyone who builds. "Be builders. Build something better," he said.

I came away from the stream still holding a box of burned DVDs from a ceremony that, by Diego's account, should never have been necessary, in a table that tells me Bitcoin sits at a zero where it matters most. The laptop and the discs are still in the box.

Listen to Diego Salazar & Luke Szramowski on Spotify, Apple Podcasts, YouTube, and More

Time stamps:

  1. 00:00:59– Introducing Diego Salazar & Luke Szramowski
  2. 00:02:04– Tribalism and Privacy Protocols in Crypto
  3. 00:03:55– Why Privacy Matters & Personal Motivations
  4. 00:04:58– Diego’s Background and Cypher Stack
  5. 00:06:58– Luke’s Background and Mathematical Perspective
  6. 00:08:40– CoinJoin and the Limits of Bitcoin Privacy
  7. 00:09:10– Threat Models and Privacy Tool Selection
  8. 00:12:22– CoinJoin’s Effectiveness and Metadata Attacks
  9. 00:14:18– Comparing CoinJoin to Tor/VPN Usage
  10. 00:17:26– Overconfidence and Under-Researched Privacy Claims
  11. 00:19:52– Levels of Privacy and User Education
  12. 00:24:31– Defining Adversaries in Privacy Research
  13. 00:26:10– CoinJoin’s Real-World Threats and Research Gaps
  14. 00:27:17– Social Media, Nuance, and Privacy Debates
  15. 00:29:11– CoinJoin as “Add-On” Privacy and Side-Channel Attacks
  16. 00:31:46– Underestimating Adversaries and Attack Tools
  17. 00:38:37– Sponsor Plugs and Criteria for Privacy Evaluation
  18. 00:42:47– Scoring CoinJoin on Privacy Criteria
  19. 00:51:09– WabiSabi and Wasabi Wallet Improvements
  20. 00:59:29– Other Bitcoin Privacy Tools: BIP47 & Silent Payments
  21. 01:06:19– Human Error and Privacy Tool Usability
  22. 01:08:51– The Limits of Privacy Ratings and the Need for Nuance
  23. 01:12:19– Tribalism, Community Research, and Productive Debate
  24. 01:15:33– Non-Interactive Proofs and Security Assumptions
  25. 01:16:27– Bitcoin Maximalism and the Limits of Debate
  26. 01:17:34– Academic Rigor and the Need for More Research
  27. 01:20:01– Bitcoin’s Privacy Risks if Mainstream Adoption Happens
  28. 01:21:44– The State of Bitcoin Privacy Research and Funding
  29. 01:34:03– Bitcoin’s Missed Privacy Opportunities
  30. 01:37:14– Monero’s History and Privacy Technologies
  31. 01:41:03– Monero’s Anonymity Set and Privacy Trade-Offs
  32. 01:46:52– Monero’s Current Weaknesses and Upcoming FCMP
  33. 01:55:24– Monero’s Privacy Scores and Scalability
  34. 02:27:50– Zcash: History, Community, and Technology
  35. 03:00:29– Zcash’s Privacy Model and Accumulator Approach
  36. 03:05:56– Zero-Knowledge Proofs Explained
  37. 03:20:08– Zcash’s Privacy Scores and Tachyon Scalability
  38. 03:35:12– Firo (formerly Zcoin) and Spark Protocol
  39. 03:44:45– Firo’s Privacy Scores and Future Upgrades
  40. 03:52:06– Litecoin MWEB and Mimblewimble
  41. 04:01:32– MWEB Privacy Scores and Limitations
  42. 04:07:03– Other Privacy Projects: Zano, Beam, Salvium
  43. 04:10:02– Toxicity, Community, and the Human Element
  44. 04:17:00– Philosophy of Privacy and Human Nature
  45. 04:24:28– Fungibility, Bitcoin, and Social Realities
  46. 04:44:15– Summary and Chart of Privacy Scores
  47. 04:57:52– Resources for Privacy Research and Final Thoughts

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
290616193d6cbc00904f2947255d36be798e3d0cb65c2d83b989ef11f0a5bd85
Proof
2025-07-03_s16_e32_bitcoin_vs_monero_vs_zcash_with_cypher_stack_s_diego_salazar_luke_szramowski.mp3.ots · all episode hashes (proof)
Status
Anchored in Bitcoin block 974,312
Timestamped
After release. The episode came out on 3 Jul 2025, but this proof was made on 3 Nov 2026 and anchored in block 974,312, so it proves this exact file existed by 3 Nov 2026, not on the release date.
How to check it yourself
  1. Download the MP3 and the proof into the same folder.
  2. Run sha256sum 2025-07-03_s16_e32_bitcoin_vs_monero_vs_zcash_with_cypher_stack_s_diego_salazar_luke_szramowski.mp3 (on a Mac: shasum -a 256 2025-07-03_s16_e32_bitcoin_vs_monero_vs_zcash_with_cypher_stack_s_diego_salazar_luke_szramowski.mp3). The result must match the SHA-256 above.
  3. Install the OpenTimestamps client (pip install opentimestamps-client) and run ots verify 2025-07-03_s16_e32_bitcoin_vs_monero_vs_zcash_with_cypher_stack_s_diego_salazar_luke_szramowski.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 2025-07-03_s16_e32_bitcoin_vs_monero_vs_zcash_with_cypher_stack_s_diego_salazar_luke_szramowski.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 trackers, no tracking cookies.

0:000:00