Even if your private key leaks, you can prove you're the rightful owner — from your seed phrase.
A full breakdown of Cardano's 7-hour X Space

The big picture first

🗺The full map of 7 hours

It's a long haul, so here's what was discussed and when, in chronological order. The color coding is a rough guide to how "dense" each part is.

Just the conclusions from the tech & strategy parts

In three lines (the core)

Who was talking

Main speakers

A star-studded session centered on Cardano, with key people from other ecosystems dropping in too.

* Since no automatic speaker separation (diarization) was performed, attributions of who said what include inferences from context.

Topic · Charles's latest news (recording ~1:53)

🏥Hoskinson Clinic handed over to Monument (a Mayo-affiliated group, per Charles)

The first thing Charles talked about when he showed up wasn't crypto at all, but a personal piece of news: the transfer of his own clinic.

Charles then went on to declare this year's focus — the "big four" — which we'll look at in detail in the next section.

Topic · Charles's priorities (recording ~1:55, and throughout)

🎯Charles's "Big Four"

The four ventures Charles named as this year's priorities are Midnight / Midnight City / RealFi / Pogun. On top of that, a meta-goal: "get Cardano's governance where it needs to be." IOG is transitioning toward a venture-studio model, and he said "we want to ship two or three ventures a year" and "we don't need many wins — three to five and Cardano grows organically."

Topic 1 · The centerpiece of this Space [Feature]

🔐A browser-based ZK "wallet ownership proof" app

The first practical ZK (zero-knowledge proof) app on Cardano. Without ever revealing your private key, you can prove "this wallet is mine" straight from your seed phrase (24 words). As Phil put it: "As far as I know, this is the first meaningful practical use of ZK on Cardano. I'm quite proud of it."

What the app does

• Goal: even if your private key has leaked, prove ownership from your seed phrase (master private key) and recover the funds.
• How it works: it generates a zero-knowledge proof that "I hold the master private key, and it derives to the credential in question." Neither the public key nor any signature is revealed.
• Status: already live on pre-prod (testnet). You can run the whole flow on your own PC — lock funds, generate a proof, and claim them.
• Ripple effects: once you can do Groth16 verification on-chain, applications open up all at once — bridge designs, reusable DeFi patterns, and more. "This isn't a small thing; it opens up an enormous set of use cases" (Charles).

Why only the rightful owner can recover "even if the key leaks" (FAQ)

"If the private key leaked, wouldn't the seed phrase leak too?" — a question many people have. The answer is No. That's because the relationship is one-way. Intuition points the "arrow" in the wrong direction.

Seed phrase (24 words)   ← root = master private key
      │  one-way transform (hashing)
      ▼
  Account key
      │
      ▼
 Signing key for each address   ← leaf = the key that signs transactions
      │
      ▼
  Public key / Address

Performance specs (for K21)

What does K21 mean? (compared with K20)

K is a metric for the size of the circuit, meaning 2ᴷ "constraints" (a constraint = a fine condition that pins down the correctness of the computation). Each increment of K roughly doubles the circuit. So K21 is about twice the size of K20. Rather than "comparing against something," the relationship is: you could "lower" it to K20, but they deliberately keep it at K21.

💡 Broken down — even though lowering to K20 would make it lighter, to prove a "stronger, safer-side claim (knows the master key)," they deliberately chose the heavier K21. Even at twice the proving weight, they prioritized a security margin (Charles: "I just ran it at K21 on purpose").

Tech stack

Topic 1-B · One of this Space's standout technical highlights

🧩Has the Gnark × WASM "4GB wall" problem been solved?

First, the gist: building this ZK proof requires a huge amount of memory. But the memory usable in a browser (WASM) is capped at about 4GB. The prover(= the software that produces the proof. The counterpart to the "verifier.") wants more than that, so as-is it won't run in a browser — this is the "4GB wall." The story is how it was overcome with "a clever way of splitting things up."

What exactly was the problem

1. A design that assumes 32GB vs. the 4GB WASM limit. The original prover was designed to write the proof into 32GB of memory. But WebAssembly (wasm32) limits linear memory to about 4GB. K21's proving key plus the MSM/FFT working area don't fit in 4GB.
2. Attempting sharding. To fit the 4GB constraint, a sharding setup using Web Workers was built. But "I noticed there were a few problems."
3. Patching Gnark itself. So Charles patched Gnark (pronounced "geosnark" in the Space) directly, arriving at a form of "fix it this way and it's solved." "It took a little while to find the way to solve it — it was probably around 1:30 in the morning on Saturday."
4. The punchline that surfaced later. It turned out that on Gnark's GitHub there was a "mega-thread" where people had been struggling with this WASM problem for years. Charles's patch was a high-quality solution that already achieved what that big effort had been aiming for — and he'd done it without knowing, just solving it within his limited time.

Topic 2 · Why this technology was needed

🏭The "SecondFi" incident, the presence or absence of KYC, and "the people who are hard to rescue (the unhappy path)"

This ZK app was born to return, to their rightful owners, funds whose keys leaked in the SecondFi wallet incident. But the story branches into a "happy path" and an "unhappy path."

Background: the "redemption sweep" of the original ADA sale (Attain) (= separate from SecondFi)

• The trigger: the seller (Attain) went out of business. Personal data may have leaked / been taken by an employee, raising the concern that at-rest secrets capable of restoring redemption accounts could be used by an attacker.
• The situation: because it was managed with multisig (multiple people had access), commercially it was at the stage where "redemption is not yet complete." The final-stage transfer from the seller to the user's sole-control wallet remained.
• The response: the funds were moved to "a different warehouse (a new redemption destination)," and users were asked to update where they collect from.
• The grace period: a mechanism was also set up so that if you don't come to collect within a set period (10 years), the transaction becomes void and you are refunded.

The happy path and the unhappy path

Topic 3 · Common Misconceptions

❄️The Glacier Drop is a "passive process"

Topic 4 · Cross-Chain Collaboration

⚛️Post-quantum crypto × Algorand — Bruno drops in

Algorand Foundation CTO Bruno Martins (formerly at IOG, until around 2021–22) dropped in. Centered on lattice cryptography, he spoke passionately about joint development that spans ecosystems. There was even a joke: "We send our good IOG engineers to Algorand — John Woods was one of them."

The "stack" they want to build together

• Post-quantum VRF: Silvio Micali is the originator of VRF + cryptographic sortition. A post-quantum version of it, built jointly. Chris may publish a paper early next year.
• Lattice-based folding: Building on a paper by Dan Boneh and Microsoft Research (Srinath), IOG is on the verge of achieving ultra-fast folding, GPU-optimized with tensor CCS + CUDA.
• Proof embedding (outer proofs): A proposal to embed the outer proof with Falcon. Dan Boneh sees room for lattice proof sizes to shrink by another half to a quarter, putting 10–20KB proofs within reach.
• Formal verification: A Rust reference implementation plus Agda / Lean blueprints, forming a multi-institution "canonical library." Drawing on lessons from the Everest project, which verified TLS in F*.
• Foundation: latticefold is already implemented in Rust under the Linux Foundation's Nightstream. A consortium model that also pulls in Stanford and others. Formalizing math problems like module SIS is also under consideration.

Topic 5 · Investment Strategy

🚀The Orion Fund and "3–5 unicorns"

Applications for the Draper-affiliated Orion Fund closed on this day. The discussion: it could become the "hub of venture-building" for Cardano. The Orion Fund / Comeda Labs (a venture studio) / Draper Dragon / Draper University form a single universe, aiming to attract large amounts of outside capital.

Topic · A listener-driven blowout (recording 4:00 / 5:30)

🦄The "Unicorn Reinforcement Loop" Debate

Following Charles's remark that "we need 3 to 5 unicorns," one listener (nicknamed Futurist) made an impassioned case for "manufacturing unicorns by booking revenue back and forth between companies." A finance pro fired back that "that's effectively fraud," making for one of this Space's most heated exchanges.

The conversation branched out to the SpaceX IPO (96% lockup, roughly $2T market cap immediately after), acquisition strategies of payment companies, on-chain analytics businesses like TRM Labs, and even a space-mining fund called "M31." One listener argued that "Jobs and Musk, aside from SpaceX, are all B2C. Today's consumers are hard to convince, so the real battle is a B2B/enterprise mutual-reinforcement loop," closing by saying he wants to put the question to Charles himself next time.

Topic 6 · The heated main event

💡What "Marketing" Really Is: "Vision"

Charles's long answer to the community's call (from Alpha) for "more marketing."

STEP 1 — The shakeup: firing back with "which matters more?"(= deliberately not giving an answer)

👉 And that's exactly why Charles's "answer" is this: every one of these points has "a legitimate difference of opinion." People who say "do marketing" often just want someone to broadcast a "specific answer" already in their head. So first get community agreement through governance on "the core KPIs and the 5-year goal" — that, he says, is the precondition for effective marketing.

STEP 2 — Why "vision"? (four pillars supporting the conclusion)

Charles backs up the conclusion above ("marketing = vision") with these four points (= his reasons for "why vision before advertising"):

• ① Start with the end in mind: assume you've closed a big round three years out, then work backwards to lay out milestones and KPIs. → In other words, the "goal (vision)" comes first.
• ② "Me too" loses: "we also have high TPS / smart-contract support" isn't differentiation and leads to a race to the bottom on price (e.g., SUI is faster and better than Solana yet still can't win on adoption). → So you need a "different vision."
• ③ Vision attracts the doers: "execution is downstream of vision. Great executors can be hired anytime. What's rare is a vision people genuinely want to join." → With a vision, the people come.
• ④ Back to roots: make Cardano an identity and a movement, more than "just a currency." The legacy built over 10 years, decentralized governance, EUTXO, Ouroboros — you can't buy or replicate those. → This is the very core of the vision worth telling.

Topic · App spotlight [Feature] (recording ~3:40 / 3:52-3:59)

🏦RealFi Feature — A Deep Dive Into "The ETF of DeFi"

The app Phil called "what I'm most excited about" and Charles rated as "an absolute delight." It's a passive financial vehicle that requires no active position management. Let's break down the key points in real detail.

① What was the problem in the first place (traditional Cardano DeFi)

Boiling down Phil's explanation — earning yield in today's DeFi takes this much effort:

1. Prepare two tokens: for example, ADA and a "Foo token," and pair them up.
2. Create LP tokens: deposit both into a DEX to provide liquidity → receive LP tokens.
3. Staking/farming: deposit those LP tokens into another contract and earn rewards in farming tokens.
4. ⚠️ Here comes Impermanent Loss: if the two assets' prices diverge, you take a loss. "It's called 'impermanent' only because semantically it isn't finalized yet — in reality it's a bona fide, permanent 'real loss'" (Phil).
5. Constant maintenance required: watch the charts, rebalance, and keep shifting liquidity between pools as volume changes.

② RealFi's answer = "The ETF of DeFi"

③ How it's actually run (the internals)

To the question "Does an AI agent run it?", Phil answered that it's run via two allow lists.

④ RealFi's "story," as Charles sees it

Charles frames RealFi as an investment narrative like this — it's a microfinance play whose yield source sits outside of crypto, so "it's a good product that works even in a bear market" and "it has the lens of banking for the unbanked, with a feel unlike anything else." He's pinning hopes on it as one of the 3-to-5 unicorn candidates.

Topic 8 · Designing governance

🏛Governance and the "executive function"

The idea of "moving the conversation off X"

Once it becomes a contest of appearances, debate breaks down. X — which runs on grandstanding by people who just want to be seen, and on rage bait that stokes anger — is fundamentally ill-suited to "consensus-building." That is the concern.

💡 Here's the crux: the idea is to build "a Cardano-only conversation channel that only ADA holders can enter." The foundation already exists — Midnight's Discord has roughly 49,000 members, they have built in-house tools comparable to the Ethereum world's Guild.xyz / collab.land, and they want to launch it in 3–6 months. The aim is to move debate from "toxic public social media" to a "deliberative space" backed by the verified identity of being a "holder."

Also drawing attention is a translation gap around Emurgo's "departure." In English it was "stepping away from the Pentad," but in Japanese it was rendered as "stepping down from being a founding entity of Cardano" — so the note goes (*mentioned within the Space; for the official position, check each company's primary sources). IOG itself is transitioning to a venture-studio model (an organizational form that continually spins up and grows new businesses in-house), and the Haskell node is slated to be spun out to a different steward next year.

Topic · Deep dive (recording ~2:15 / ~3:05, and various other points)

🌙Midnight deep dive — a new kind of ID through "selective disclosure"

Midnight, which Charles named at the top of the "Big 4." Here we organize the key points raised in the Space, digging into the technical substance. First, as a foundation: Midnight is a privacy-focused partner chain, a separate company with its own blockchain and its own budget (kept separate from Cardano's Treasury).

The core = ID through selective disclosure (what Charles calls "Midnight Passport")

*"Midnight Passport" is the name Charles used in the Space; we have not been able to confirm it as an official product name. That said, the selective-disclosure ID capability itself does exist in Midnight (see the note at the end).

Where it's useful (applications raised in this Space)

• The challenge exposed by the SecondFi incident: the idea of "proving you're the owner after the fact even if your keys leak" — the context in which Charles said "that's why we need this kind of ID" (→ SecondFi incident). *The actual recovery app this time was not Midnight but built with Cardano/Gnark.
• Safe post-quantum migration: because you can demonstrate ownership without relying on signatures or public keys, it can be used for migrating to post-quantum addresses (→ see the explanation at the end of Gnark×WASM).
• Anonymous voting: enabling people to vote in governance without fear of retaliation (see below).

Anonymous voting and governance

Charles points out that when people fear retaliation or bullying, they start to "stop voting / only act when it's in their own interest," giving rise to the 19th-century American Tammany Hall-style quid pro quo voting. As a countermeasure, there was talk of a plan to put part of Catalyst on Midnight and introduce anonymous voting.

Community and tools

Topic · Practical Q&A (recording 5:00–6:00)

🌙Midnight's partnerships, budget, and ambassador activity

In the latter half, questions from the community came thick and fast about the organizational and partnership realities around Midnight.

Topic · Actually one of the biggest chunks (recording 0:00–1:00 / 5:00–6:00, etc.)

🤖Behind-the-scenes AI industry talk — Anthropic vs OpenAI

It's a Cardano Space, but in fact candid assessments of AI coding tools and each company's models took up a long stretch. We summarize the key points as the panelists' "voices from the field" grounded in their own experience (this is strictly personal opinion and gossip, and accuracy is not guaranteed).

Topic · Practical material (recording 6:00–6:40)

🔧"Practical wallet tips" picked up from the chatter

Mixed in with the late-night chatter, there was actually some useful knowledge too. It's based on the panelists' anecdotes, so act at your own risk and after verifying.

Topic · Other moments

🎬What happened in the Space (digest)

Supplement

📖Mini glossary (clarity first)

Hard words explained with everyday analogies. Want to dig deeper? Follow the "🔗 Learn more" link on each card.

Primary source

📄Original transcript (full 7-hour version)

This summary is an unofficial reconstruction by a third party, based on the full text (about 65,000 words) of the roughly 7 hours 13 minutes of X Space audio (title: Game Time 🫶🏽 (Unification station)), which was auto-transcribed by AI (Whisper large-v3-turbo). For those who want to "throw it at their own AI to summarize" or "check the primary source," we've made it possible to view, copy, and download the full text.

⚠️ Because it's auto-transcribed, it contains mishearings, inconsistent spelling of proper nouns, and speaker mix-ups. No speaker separation (diarization) was performed. The crypto-tech discussions covered on the summary page appear after roughly 2:00:00 in this recording (the first half is chit-chat and other topics).

🔗 Original audio: x.com/i/spaces/1rGmqqEAmZqGy — recordings of X Spaces can become unavailable after a certain period. That's exactly why this full transcript (below) has been preserved.

Open "Show full text" to load the transcript here…