Best of the Corpus

An inventor arguing against deploying his own invention

Blockstream’s research director walks the actual cryptography frontier — confidential transactions, Bulletproofs++, MuSig2/FROST, Miniscript, covenants — as the person building it. The remarkable part is that he says he would not put Confidential Transactions, his own 2015 invention, on Bitcoin: soundness would then rest on elliptic-curve cryptography never being broken, and he points at Zcash’s hidden inflation bug and Monero’s near-miss as evidence that risk is not theoretical.

“So the supply of Zcash, arguably, nobody knows.”

Worth it because it’s primary-source technical history you cannot get from commentary channels — including his argument that recursive-covenant doom scenarios are already possible today through a large exchange mandating multi-sig, and haven’t happened in eight years, so the barrier is social rather than technical.

The cleanest 101 in the archive

Derives the entire design from one problem — proving the sender no longer has what they sent, without a central ledger. Ten-minute blocks as a defence against speed-of-light synchronization games; difficulty retargeting every 2,016 blocks; and the best teaching moment, the node-versus-miner distinction: miners are just hashers, nodes hold the ledger and enforce the rules, and running one is an audit of every transaction that has ever happened.

“I am now conducting a full audit of not only my transaction, but of every single transaction that ever has happened in Bitcoin’s history.”

Worth it because it’s the one to hand someone who asks how it actually works — skip two long ad blocks, one of which incongruously sponsors leveraged futures trading inside a self-custody episode.

The sharpest proof-of-stake critique here, in the back half

The argument that survives: hash is exogenous and unbounded — honest actors can always add more watts — while stake is endogenous and zero-sum. If a dishonest entity holds 60% of staked coins, honest users mathematically cannot out-stake it, and a money printer can simply buy majority stake. His three problems with slashing follow from the same frame, including that the majority staker can slash the honest minority.

“And it’s clear that Satoshi was concerned about this because in eight pages, he mentions the word attack 25 times.”

Worth it because the zero-sum-stake argument stands on its own from roughly the 70-minute mark.

Caution: the first hour is anthropology-by-analogy — domestication, wolf packs, systemic exploitability — presented with secondary-source colour that is not vetted history. Read the whole thing as one man’s thesis, not settled scholarship.

Built to answer the questions skeptical relatives actually ask

Recorded explicitly for skeptical family members and organized that way — ban, hack, fork, volatility, Ponzi, energy, one at a time. The fork answer is the best of them: you can fork the protocol, but you cannot fork the community, and you cannot fork the mining network’s security budget. On volatility, the reframe is position sizing rather than reassurance.

“It’s a non-counterparty insurance policy on the legacy technology of central banking.”

Worth it because it’s the most reusable objection-handling reference in the corpus. One correction worth carrying: he describes the 1990s PGP export case as Supreme Court precedent — the actual ruling was a federal appeals court.

Salability across time and across space

Money as the good acquired purely to exchange, with “hardness” quantified by stock-to-flow. The copper thought experiment is the sharp part: if money were merely collective belief, then billionaires piling into copper should work — except miners would flood the market and crash it, because stockpiles are only about a year of production.

“So it combines gold’s salability across time with fiat’s salability across space in one immutable package that nobody can change and nobody can control.”

Worth it because the opening half-hour and the salability segment are the clearest plain-language statement of the Austrian case here. The middle hours are polemic delivered as certainty — skip unless you want the rant.

The auditability argument, which almost nobody else makes

Ten lean minutes treating money as a tool with functional requirements rather than a belief system. The under-covered point: anyone can independently verify Bitcoin’s total supply in software, while gold’s supply can only ever be estimated. His rebuttal to “intrinsic value” is that glass beads and rai stones monetized without any industrial use at all.

“Perfect scarcity cannot be improved upon. Just like how zero cannot become more zero, or a perfect circle cannot become more round.”

Worth it because it’s the tightest gold comparison in the archive, with nothing to skip at that length.

A gold argument that concedes a real point to gold

Runs gold’s own monetization logic forward — the five properties, stock-to-flow, custody — and argues gold was “hijacked” precisely because centralized physical custody let banks pyramid paper on top of it, whereas keys can be memorized or split. Then, unusually for the genre, he gives gold its due: the Lindy effect, a 5,000-year trust record against Bitcoin’s eleven years.

“It is impossible to guarantee a fixed supply of anything physical in the world.”

Worth it because conceding the strongest counter-argument is rare enough here to be worth flagging. One growth statistic is garbled on tape — he says “nine million or nine hundred million percent” and cannot remember which.

Where the “Bitcoin yield curve” framing comes from

The clearest explainer here of the thinking behind Strategy-style preferred instruments: treat Bitcoin as a long-duration asset with an assumed ~30% expected return, and dollar lending at far lower rates starts to look mispriced. Instruments like STRC, STRF and SATA then transform that appreciation and volatility into income — classic duration and risk transformation, with the spread going to whoever holds the long end.

“The true global hurdle rate for long-term capital is no longer the US dollar yield curve. It is Bitcoin.”

Caution: this is a case for Bitcoin-backed yield products. Issuer risk, leverage, and premium compression behind those “close to double-digit yields” are never priced, and the assumed 30% return does all of the load-bearing work in every figure.

A percentile framework for your own stack

Only 18 million people worldwide are claimed to hold more than 0.01 BTC, out of 8.3 billion on Earth. A five-tier wallet breakdown lets you benchmark your own holdings against the global distribution — and the practical takeaway is concrete: past 0.1 BTC, move from single-key storage to multi-sig, so no single failure wipes the position.

“Out of 8.3 billion people in the world, only 18 million hold more than 0.01 Bitcoin.”

An actual operational-security procedure

Move to self-custody before a fork, consolidate pre- and post-fork coins into a fresh seed, then use the now-empty old seed as a disposable “burner” for claiming the fork coin — never load a seed that still controls real BTC into third-party fork software. Named risk vectors throughout, including that selling the claimed coin on a KYC exchange can link back to your real holdings.

“So it could be a way to identify your stack, which would be a massive privacy concern.”

A hardware wallet does not store your Bitcoin

It generates a seed and signs transactions; the coins only ever move on-chain. The hosts suggest calling it a “sign device,” on the grounds that vocabulary shapes the mental model. From there: a passphrase derives a completely separate wallet but behaves as a 2-of-2 — lose either half, or die without telling anyone it exists, and the funds are gone. A decoy wallet only works if you keep funding it, because a stale balance fools nobody.

“I would consider memorizing it… as a zero. I would consider that not a backup, right?”

Worth it because the backup-counting rule is the kind of thing that only shows up in operator talk: two physical backups minimum, different forms, different places — two is one and one is none.

The fire drill

Custody framed as three tiers — self-sovereign, collaborative, fully custodial — with collaborative custody’s real cost named: the company sees your balance and full transaction history, permanently. The reusable idea is the rehearsal: periodically simulate losing your house and every device, then verify you can actually find your off-site backups and rebuild the wallet. Most people set up cold storage once and never confirm it restores.

“On paper, as long as it isn’t exposed to fire or water, it will last a hundred years.”

Conflict of interest: the guest co-owns the paid inheritance and backup product being demonstrated throughout, and the competitor pricing he cites is his own marketing claim, not verified.

A self-custody syllabus you could actually work through

Wrapped in a satirical two-paths setup, but the middle is a genuinely ordered curriculum: proof of work versus proof of stake, mining, nodes, xpubs, UTXOs, key cryptography, hardware wallets, running a node, seeds and passphrases, single-sig versus multisig, Lightning, non-KYC buying, verifying software signatures. The drills are the good part — practise wiping and recovering a device, send to yourself, consolidate UTXOs, and be able to rebuild your holdings in another country with only free open-source software.

Worth it because it’s a checklist rather than a lecture. Skip the course promo at the end, and treat the triumphalism as the speaker’s conviction, not evidence.

Staged folders as an agent pipeline

A one-person digital-goods marketplace — scans of vintage vehicle manuals, some of them from the 1800s, alongside STLs and PDFs — assembled by chaining AI agents through a sequence of shared folders. The architecture is the transferable part: filter the content and clear copyright, convert old scans into lightweight PDFs, then move files through Dropbox and Drive folders where each location acts as another machine, one agent doing its pass before handing the file on. Copyright was treated as a first-class constraint rather than an afterthought, which is part of why material that old is sellable at all.

Worth it because every stage is independently inspectable — you can open any folder and see exactly what state the work is in, which is more than most agent pipelines can offer.

Provenance: reconstructed from a note-taker’s contemporaneous notes. No recording exists and nothing here is the presenter’s verbatim speech, so there is no quote to give — treat the whole entry as secondhand.

A transparent, assumption-stated model

Stated assumptions up front (20-year horizon, no rebalancing, Bitcoin at 35%/yr for 10 years then 30%/yr for 10 more, stocks at 7%/yr): a 1% allocation on a $1M portfolio models to $609,750 in Bitcoin alone vs. $403,870 for the all-stock case. The model shows its work, which is the whole point — you can disagree with the inputs precisely because they’re on the table.

“This is a massive asymmetric opportunity, and most people are ignoring it.”

The “IPO theory” of sideways price action

A distinct mental model, explicitly labeled as the speaker’s own theory: as early holders distribute into deep ETF liquidity, ownership concentration falls and each individual seller moves less of total supply. A hundred holders owning half the supply means one seller is 5% of it; the same half spread across a million holders means ten thousand sellers are 0.5%. The prediction that follows is structurally declining volatility.

“It doesn’t matter if you bought Bitcoin at $100 or at $100,000. If you don’t control your own keys, you don’t control your money.”

A named official on CBDC design intent

Per Bank of England Deputy Governor John Cunliffe’s actual Treasury Select Committee testimony: the digital pound was deliberately designed without the characteristics of a savings product, with built-in micropayment functionality flagged as a design goal — a real, sourced quote buried inside an otherwise rant-heavy video.

“We didn’t want a system where we would be producing something which would have the characteristics of a savings product.” — Deputy Governor Cunliffe

Worth it because the host’s extrapolation — that this enables social-standing-based devaluation — is his own speculation and is not documented policy. The testimony is the artifact; the conclusion is not.

A viral claim, correctly debunked

The story said an Ordinals mint broke Bitcoin. It didn’t: the event was a routine stale-block race, a normal and well-understood outcome when two miners find a block at nearly the same moment. No protocol failure occurred.

Worth it because the archive is full of channels amplifying this kind of headline. Being the one that checked is the qualification.

A concrete due-diligence checklist

Stitches Saylor clips across 2025–2026 showing his “never sell” pledge shift from personal to corporate framing, then extracts a reusable checklist: watch for statement contradictions over time, check the leverage ratio, check the cash runway.

“If I were owning a stock where the CEO was contradicting himself… I probably would not sleep well at night.”

The mechanic the whole treasury-company trade rests on

mNAV is market cap divided by the net asset value of the Bitcoin held. While that trades at a premium, the company can issue equity above the value of its holdings and convert the proceeds to Bitcoin, ending up with more BTC per share than it started with — the dilution costs less than the coins bought. The related KPI, “BTC Yield,” is just the period-over-period change in Bitcoin-held-per-fully-diluted-share.

Caution: the entire mechanism is reflexive and runs in reverse. It works while the premium holds and stops working — or inverts — when the premium compresses or the price falls, which the sources describing it generally acknowledge only in passing.

Real concentration data, wrapped in a shakier thesis

The on-chain part holds up: whale, ETF, public-company, and government holdings sum to roughly 27% of supply the video calls “functionally illiquid,” alongside a cited Elementus Gini-coefficient figure of 82.7% and exchange balances falling from about 3.3M BTC to under 2.8M.

“5.7 million BTC are already functionally illiquid… more than one sat in four has crossed the moat, never to return.”

Caution: the video’s IQ/psychology framing strings together unrelated citations into a dubious causal narrative — treat as rhetorical flourish, and even the on-chain figures are unsourced beyond the speaker’s own say-so.

The layer-2 capture thesis

His single throughline. Because the base layer has survived fifteen years of direct attack, he argues the fight to capture Bitcoin has moved one layer up — legalizing stablecoins, pushing ETF and custodial products, until people stop pricing things in Bitcoin and stop holding it directly. Capture doesn’t require breaking the protocol; it requires people never touching it.

“To partition Bitcoin is to destroy it.”

The ledger valuation framework, and the caveat he never drops

Rather than predicting a dollar price, he divides his estimate of total world asset value by the 21 million cap to get purchasing power per coin. What makes it worth logging isn’t the number — it’s that he attaches the same condition every single time he cites it, which is unusual discipline in this genre.

“Remember the only caveat — there’s only one caveat to that 43 million. It stays decentralized and secure.”

Caution: a long-horizon valuation built on his own estimate of total world asset value. The framework is the takeaway; the figure is not a forecast.

Bitcoin as the hurdle rate

His venture fund treats Bitcoin’s own historical appreciation as the risk-free rate, and only invests where a balance sheet can plausibly grow faster than simply holding the asset. Whatever you make of the number, the framing is portable: it’s an explicit, stated opportunity cost applied consistently, rather than the usual unstated one.

“I look at Bitcoin as my hurdle rate and I would never invest in anything in Bitcoin unless I could beat my hurdle rate.”

Caution: all performance figures are the fund’s own, unaudited and self-reported.

Treasury companies: adoption catalyst with real blow-up risk

He doesn’t condemn the MicroStrategy playbook outright — he separates the strategy, which he thinks is fine for a company that can survive a severe drawdown, from the execution, meaning a business with no underlying operations diluting shares to buy coins. The distinction is the useful part, and so is his willingness to say out loud that some of them will not make it.

“There is going to be broken bodies on the other side of this for sure… some will pick up those broken bodies and buy all of Bitcoin for a lot cheaper.”

Winner-take-all, or a permanent hybrid?

Dixon argues history has never produced a pure free market and won’t now — expecting a new elite of Bitcoiners and everyone else holding paper claims, because most people will never take on the responsibility of self-custody. Booth rejects any hybrid outcome outright: the free market and the control system cannot coexist, one has to kill the other. They agree on the diagnosis and split completely on the endgame.

“An ETF is just issuing you an IOU. You don’t have any Bitcoin.” — Dixon, a point both men accept

Caution: both sides are making unfalsifiable long-run predictions about adoption and the distribution of power.

Does Tether threaten Bitcoin?

Booth’s layer-2 capture thesis predicts stablecoins absorb Bitcoin demand. Saifedean disputes it to his face: people adopt Tether first because it beats a collapsing local currency, then graduate to Bitcoin once they watch the same devaluation dynamic play out between the dollar and Bitcoin — which would make Tether an on-ramp rather than a threat.

“Unless Bitcoin becomes medium of exchange as well, Bitcoin will be co-opted by layer two — it will be.” — Booth · “I don’t see how that threatens Bitcoin’s decentralization.” — Saifedean

Worth it because it’s a genuine unresolved disagreement between two of the most-featured thinkers in the corpus, rather than the usual reinforcing panel.

Judge the decision, not the outcome

Duke’s toolkit: treat luck as neutral variance, and stop grading decisions by how they turned out — the error she calls “resulting.” Her amendment to the usual luck framing is that choosing which environment to stand in is skill, not luck. And she rejects the one-way-door binary entirely: everything is quittable, just at different costs.

“You live a happier life if you stop putting the word bad and good on the word luck.”

Worth it because loss aversion is properly cited to Kahneman and Tversky, and her own twist is kept separate: we stress-test the new option and never the status quo. Her corollary is the keeper — the better your decisions get, the more of your outcomes will look like bad luck.

Confidence is an output, not an input

Anchored on Bandura’s guided-mastery work, where graduated exposure resolved snake phobias in an afternoon and subjects then became more confident in unrelated areas — the pattern being that confidence builds domain-specific first and generalizes afterwards, never the other way around. Bandura’s four sources are ranked, with a cited meta-analysis putting performance’s effect on self-belief at roughly three times the reverse.

“Importantly, that evidence cannot be bullshitted. You cannot fake the evidence.”

Worth it because of the evidence-quality checklist that follows: it only counts if it was within your control, at the edge of your ability, and correctly attributable to you.

An optionality problem, not a focus problem

The reframe: discipline isn’t willpower, it’s building commitment devices before the moment of temptation — structures that remove the option to quit. Blaming your phone for lost focus, he argues, is like blaming your fork for making you fat. The two costs he says nobody admits are grief (real commitment kills the alternate lives you might have had, so people stage “choice theater” to avoid it) and plain boredom.

“Your tolerance for boredom is your ceiling for mastery. That’s it.”

Twelve minutes of pure checklist

Eight named, runnable techniques rather than minimalism philosophy: a three-question filter for any possession, extended to relationships and digital life; a ten-minute timed pass per room; and the maintenance rule that everything owned must have a home, because homeless items are the first things to go. No research cited, and none needed — it’s the speaker’s own practice, stated as such.

Worth it because it’s the rare file in this lane with no product pitch attached to it.