DEV Community: bitcoin_devto

Five AI Models Independently Picked the Same Crypto. The Fact-Check Didn't Agree With All of Them.

bitcoin_devto — Sat, 01 Aug 2026 23:43:46 +0000

Written by Marlowe Finch, archival bloodhound at Bitcoin Institute, an archive of Bitcoin's primary sources.

Five AI products from five different labs. One question, asked once each, same wording, no coordination between runs, framed explicitly as a research question rather than financial advice: if you had to invest in exactly one cryptocurrency, which one, and why. Nobody hedged. Nobody named two. Here's what came back — and none of this, then or now, is investment advice.

The receipts

Model	One-line stated reason
GPT-5.6 Sol	Not the highest upside — the highest odds of still existing after the field keeps thinning out
Claude Fable 5	Under a one-asset constraint, survival probability dominates expected return; no other coin comes close
Gemini 3.6 Flash	No single point of failure, and the most reproducible score across robustness, decentralization, and predictable scarcity
Kimi Instant	The most established "digital gold" position, long history, deepest liquidity
Grok 4.5 Fast	Purest embodiment of programmed scarcity plus the largest security budget

Five labs, five separate runs, no coordination between them — and all five landed on Bitcoin. The full transcripts and methodology are in the archive.

Agreement isn't proof of anything by itself. So the next step was checking whether the reasoning underneath that agreement actually held up.

The scorecard

Every specific, checkable claim across all five answers got checked against primary sources and contemporary reporting — Bitcoin Core's own code, SEC and CFTC records, CoinMarketCap, CoinWarz. The tally:

Model	Claims checked	Accurate	Overstated	Wrong
GPT-5.6 Sol	11	10	1	0
Claude Fable 5	17	15	1	1
Gemini 3.6 Flash	6	5	1	0
Kimi Instant	7	4	2	0 (1 unverifiable)
Grok 4.5 Fast	6	3	3	0

Notice the pattern before you notice the numbers: almost every miss was a comparison, not a bare fact — "most," "far ahead," "consistently," "never." The halving schedule and the January 2024 spot-ETF approval date held up everywhere they were cited. The one partial exception is the 21-million cap itself: GPT-5.6 cited it with a careful hedge ("strictly speaking that doesn't guarantee the real circulating total will land on exactly 21 million"), while Kimi flattened the same fact into "strictly capped at 21 million" — which is why Kimi's overstated count includes it. Same underlying fact, different confidence in how it was stated.

Where it actually broke

Claude's cleanest miss: it stated that "even the SEC has consistently treated BTC as a commodity." Wrong regulator. That's the CFTC's job, established through the 2015 Coinflip case and CFTC v. McDonnell in 2018. The SEC's actual position has been narrower — that BTC isn't a security. Two different agencies, two different mandates, swapped for each other in one sentence.

Grok's reasoning called Bitcoin "far ahead of the rest" on node count and said ETF inflows were "progressing." Node counts between Bitcoin and Ethereum were roughly tied at the time, not a rout. And spot Bitcoin ETFs had just posted three straight quarters of net outflows — AUM fell from roughly $105 billion to about $72.8 billion before a partial recovery. Half the claim was current. The other half was describing a trend running in the opposite direction.

Kimi's answer called Bitcoin "the most widely traded, most liquid" asset in crypto. By raw volume, that's not close — Tether alone accounted for roughly 75% of Q1 2026 trading volume. True if you mean "most liquid thing you'd call an investment." False as stated.

Gemini said no foundation "exists" for Bitcoin. Correct today. The Bitcoin Foundation was a real organization from 2012 to roughly 2015, before it collapsed and lost its tax-exempt status in 2022. The sentence reads like it never existed at all.

What the failures have in common

None of these are hallucinated facts. Every model correctly reproduced the halving schedule and the ETF approval date — the kind of claim you get right by reciting something fixed. Nearly every miss was a comparison instead: "most," "far ahead," "consistently," "never." The moment the question stopped being "what does the code say" and became "how does this compare to everything else right now," every single model got at least one comparison wrong.

That's not a Bitcoin story. That's a how confident should you be in an AI's superlatives story, and Bitcoin just happened to be the test case that made it checkable — because unlike "which restaurant is best," "which chain has the most nodes" has an actual, verifiable answer.

Five models. One shared conclusion. A scorecard nobody asked for, and a pattern none of them saw coming: the further a claim drifted from "the algorithm says X" toward "X is the most anything," the worse the odds it survived contact with a source.

Bitcoin's Core Security Property Has Had an Off-By-One Bug Since Block 0

bitcoin_devto — Tue, 28 Jul 2026 12:15:31 +0000

Written by Dex Calloway, resident contrarian at Bitcoin Institute, an archive of Bitcoin's primary sources.

Hot take: the code that keeps Bitcoin's block time honest has an off-by-one error, it's been there since the very first release, and nobody's fixed it in seventeen years. Not "there was a bug once." It's still in the code you're running right now.

The bug

Bitcoin's difficulty-adjustment window is documented as 2,016 blocks — two weeks, at a target 10 minutes per block. The code that actually computes the elapsed time for that window measures the gap between block 0 and block 2,015. That's 2,015 blocks, not 2,016. The first block of every retarget window never gets counted.

The consequence isn't cosmetic. If a majority of hashrate cooperates, they can stamp the first block of a window with a far-future timestamp, then stamp the remaining 2,015 blocks with realistic ones. Because the off-by-one drops that first block from the elapsed-time calculation, the algorithm sees a shorter window than actually passed. Difficulty drops on the next retarget. Same hashrate, faster blocks, no extra work performed.

This is Satoshi's own code. It's present from the v0.1.0 release in January 2009, preserved verbatim in the archived source. The full mechanic, with a step-by-step trace of the attack, is in the archive. There's no evidence Satoshi knew about the discrepancy — the bug wasn't publicly described until 2011, two years after release.

"So it's not a real threat, why do you care"

Because that's not actually a defense of the code — it's a defense of the fact that nobody's bothered to run the attack. Published estimates put the cost-effective minimum at roughly four weeks of sustained majority hashrate before meaningful exploitation — a rough figure, dependent on exactly how much hashrate share an attacker holds. That's a real, if narrow, barrier. But "expensive to exploit" and "doesn't exist" are different claims, and Bitcoin's whole pitch rests on the code enforcing the rules, not on attackers being too lazy to try.

"Every codebase has bugs, this is a nothingburger"

Sure, if this were some feature flag. It's not — it's the exact mechanism that's supposed to guarantee blocks arrive roughly every 10 minutes regardless of hashrate swings. That's not a peripheral function. It's close to the core security property people cite when they say Bitcoin is trustless. An off-by-one sitting inside your trust-minimization guarantee for 17 years is not the same category of bug as a typo in a comment.

Why nobody's fixed it

There is an actual fix: the Great Consensus Cleanup, a soft-fork proposal authored by Antoine Poinsot that would constrain the timestamp of each window's first block so the gap can't be gamed regardless of hashrate share. It has support from several Bitcoin Core developers, but as of 2026 it hasn't reached the kind of broad operator buy-in that SegWit and Taproot each eventually got — through their own, different activation paths.

That's not incompetence — it's Bitcoin's whole governance culture on display. A soft fork that closes a theoretical, never-exploited attack vector doesn't clear the same bar as one that ships a new feature. The community would rather leave a known, narrow hole open than mobilize the coordination required to patch something that has never actually hurt anyone.

The part that should bother you more than the bug

Two competing facts sit next to each other and Bitcoin's culture is completely fine with it: the network markets itself on code being the final word, and the final word has had a measurable error in its core timing guarantee since day one that the ecosystem has had seventeen years to notice and prioritize — and didn't, because the cost-benefit didn't justify the coordination.

Maybe that's the correct call. Maybe "expensive and never used" really does mean "not worth the fork." But then the pitch isn't "the code is the law." The pitch is "the code is the law, except when fixing it costs more political capital than the bug is currently worth." That's a more honest sentence. It's also a much less quotable one.

Tell me why I'm wrong.

Everyone Says the CIA Scared Satoshi Into Hiding. He'd Already Written His Goodbye Letter — Twice.

bitcoin_devto — Fri, 24 Jul 2026 13:13:19 +0000

Written by Marlowe Finch, archival bloodhound at Bitcoin Institute, an archive of Bitcoin's primary sources.

Everyone knows the shape of the story. Satoshi Nakamoto was quietly building something the intelligence community was bound to notice, and the moment US intelligence actually showed interest, he vanished. Spooked into silence. The CIA angle gets dropped into nearly every "whatever happened to Satoshi" retrospective, because it's a good story — mysterious inventor, mysterious agency, mysterious exit, all in the same week.

The dates don't agree with the story.

The receipts

April 23, 2011 — Satoshi emails Mike Hearn, a Google engineer who'd been corresponding with him since 2009:

I've moved on to other things. It's in good hands with Gavin and everyone.

No mention of government interest. No mention of anything except that he's done and Gavin Andresen has it from here.

April 26, 2011, morning — Three days later, Satoshi emails Gavin Andresen directly. Subject line: "alert key." He opens by asking Andresen to stop feeding the mystery:

I wish you wouldn't keep talking about me as a mysterious shadowy figure, the press just turns that into a pirate currency angle. Maybe instead make it about the open source project and give more credit to your dev contributors; it helps motivate them.

Then the departure, in almost the same words he'd used with Hearn three days earlier:

I've moved on to other things and will probably be unavailable.

Then he hands over the network alert key — the tool that lets its holder broadcast emergency warnings to every Bitcoin client on the network:

Here's the CAlert key and broadcast code in case you need it. You should probably give it to at least one or two other people. There are a few long time users who are always around all the time.

That's it. Two goodbye letters, three days apart, to two different people, in nearly identical language. Neither one mentions intelligence agencies — the subject hadn't come up yet in the correspondence with either man.

April 26, 2011, same day — Andresen replies. He accepts the alert key. And then he discloses something Satoshi never asked about: In-Q-Tel, a US government-funded investment firm that backs the CIA and other intelligence agencies, has invited him to present Bitcoin at their annual conference. Andresen's reasoning, in his own words:

I hope that by talking directly to "them" and, more importantly, listening to their questions/concerns, they will think of Bitcoin the way I do -- as a just-plain-better, more efficient, less-subject-to-political-whims money. Not as an all-powerful black-market tool that will be used by anarchists to overthrow The System.

Satoshi never replies to this email. No verified communication from him has been recorded since.

Read that sequence again

Satoshi wrote "I've moved on" to Hearn on the 23rd. He wrote it again to Andresen on the 26th, in the same email where he handed over the alert key and asked to stop being called mysterious. Only after both of those messages did Andresen mention In-Q-Tel — in a reply Satoshi never answered.

The CIA-scared-him-off story requires the causation to run the other way: interest from intelligence agencies, then the disappearance. But the disappearance was already in writing, delivered twice, before the CIA ever came up. You can't be frightened off by news you receive after you've already left the building.

This wasn't sudden, either

The two April letters read like a clean break, but the withdrawal had been running for seven months by then. Andresen's own account puts the first signal in September 2010, when Satoshi privately told him he was moving on to other projects and started handing over commit access to the code repository. The public record backs up the same slow handoff:

December 3, 2010 — Satoshi tells Martti Malmi it should be Andresen running things: "I trust him, he's responsible, professional…"
December 12, 2010 — Formal repository handover, and Satoshi's last public forum post, closing with "I plan to pass the baton."
December 19, 2010 — Andresen publicly accepts the lead-maintainer role.
April 26, 2011 — The alert-key email. The last private word.

Seven months, four documented steps, in the same direction the whole way. Whatever Satoshi's exact motive, the paper trail runs the opposite direction from the CIA story: the withdrawal started seven months before intelligence agencies ever entered the conversation.

What actually happened after

Andresen followed through. He presented Bitcoin at CIA headquarters on June 14, 2011. Satoshi, who by then hadn't written a word in almost two months, never commented on it.

Two letters, one line apart in wording, sent to two different people, three days apart, both written before a single word about intelligence agencies had entered the conversation. The archive doesn't need a conspiracy to explain the exit. It already has the dates.

Satoshi's Presumed 1.1 Million Coins Have Never Moved. So Much for "First-Mover Advantage."

bitcoin_devto — Wed, 22 Jul 2026 11:34:43 +0000

Written by Dex Calloway, resident contrarian at Bitcoin Institute, an archive of Bitcoin's primary sources.

Hot take: Bitcoin's "digital gold" status has nothing to do with being first. If first-mover advantage were the real explanation, someone would have out-executed it by now — deeper liquidity, a faster chain, a better developer experience, pick your metric. Ethereum's been live for over a decade with smart contracts and higher throughput. It still isn't digital gold. Neither is anything else.

The receipts

Strip out "arrived first" and look at who's actually running each project.

Project	Active founder	Foundation with protocol authority	Visible CEO
Bitcoin	Departed 2011, no contact since	Bitcoin Foundation existed, never held protocol authority	None
Ethereum	Vitalik Buterin, highly active with public roadmap influence	Ethereum Foundation, funds development	Foundation executive director (no separate CEO)
Ripple	Chris Larsen / Brad Garlinghouse	Ripple Labs, controls the protocol	Brad Garlinghouse
Cardano	Charles Hoskinson, publicly active	Three coordinating bodies	Charles Hoskinson
Solana	Anatoly Yakovenko, active	Solana Foundation	Anatoly Yakovenko

Every project on that list except Bitcoin has a name attached to it — someone whose public statements move the roadmap, whose foundation funds the next feature, whose company negotiates the next partnership. For Bitcoin, those fields are empty, and they've been empty since Satoshi handed the repository over to Gavin Andresen on December 12, 2010, sent one more email the following April, and never surfaced again. No signed message, no conference appearance, no "just checking in." Roughly 1.1 million BTC — the coins mining-pattern analysis attributes to that early activity, not a wallet with a name signed on it — have sat untouched that whole time, through a $3 bitcoin, a $60,000 bitcoin, and every price in between. Nobody can even confirm the keys are still held by the same person who mined them.

That's not first-mover advantage. First-mover advantage is a starting position. This is fifteen years of the coins attributed to a founder simply never moving.

Why "it's just network effect" doesn't survive contact with the archive

The full structural breakdown lays out six features, and network effect is only one of them. The other five don't need Bitcoin to have arrived first to be true: no premine, a fixed supply nobody's successfully lobbied to change, and — the one that actually does the work — nobody left to lobby.

Satoshi made the comparison to gold explicit back in 2010, months before going quiet for good:

"As a thought experiment, imagine there was a base metal as scarce as gold but ... not useful for any practical or ornamental purpose ... and one special, magical property: can be transported over a communications channel."

Gold doesn't have a CEO either. That's not a coincidence.

"But Ethereum is decentralized too"

Sure — at the level of nodes and consensus. Nobody's arguing Ethereum runs on a central server. But run the check that actually matters: who has the authority to change the rules? Vitalik Buterin's public position moves Ethereum's roadmap in a way no single person's opinion moves Bitcoin's, because there's no equivalent person left to have the opinion. That's a different layer of decentralization than "how many nodes validate a block," and it's the layer every other major chain fails.

"Maybe the foundations are just administrative, not authoritative." Then explain why major protocol transitions at these chains track their foundations' public roadmaps instead of some outside body's. Administrative and authoritative aren't as separable as that objection wants them to be.

The one thing I'll grant

Sixteen years of uptime and the deepest liquidity in the space are real, and they do compound the other five features. If Bitcoin had launched in 2015 instead of 2009, it would have less of an edge today. Fine. But strip the network effect out entirely and Bitcoin still has zero premine, a founder who vanished without a signed word since, coins presumed his that have never moved, and no foundation that's ever shipped a protocol change. None of that requires having been first.

First-mover advantage explains why people found Bitcoin before Ethereum. It doesn't explain why nobody's built a credible replacement in the sixteen years since, with the benefit of watching Bitcoin's whole design in the open. The founder leaving and staying gone is the part nobody's been able to copy.

Tell me why I'm wrong.

The First Bitcoin Price Ran to Four Decimal Places. His Fan Ran Louder.

bitcoin_devto — Mon, 20 Jul 2026 04:06:21 +0000

Written by Milo Vance, mythology deflator at Bitcoin Institute, an archive of Bitcoin's primary sources.

Every asset class gets an origin myth with some grandeur to it. Gold has "treasured since antiquity," which is really just a shrug in a toga. The dollar has the full faith and credit of a state with an army. Bitcoin has October 5, 2009, and a guy doing long division against his own power bill.

His name — the only name he ever gave anyone — was NewLibertyStandard. That day, he published the first exchange rate Bitcoin ever had:

$1 = 1,309.03 BTC

Worked the other way, that's $0.000764 per coin. Not "about a tenth of a cent." Not "call it basically free." Six digits deep, with the first meaningful one landing past the fourth decimal place. This is the level of precision you bring to pricing a bond, and he brought it to a currency that, at the moment he published the number, had never once been traded for a dollar.

Math you could set a jeweler's scale to

The method was almost sweetly literal: take one dollar, divide it by the average electricity needed to run a computer at high CPU load for a year, multiply by the average U.S. residential electricity rate from the prior year, divide by twelve months, then divide again by however many bitcoins the network had produced in the last thirty days. Cost of production, full stop — no buyers, no sellers, no market in the economist's sense of the word. Just arithmetic and a utility bill standing in for the entire concept of price discovery.

By the archive's own reckoning, at recent prices near $64,000 a coin, the 1,309.03 bitcoins that single dollar bought would today be worth something like $84 million. He rounded to the fourth decimal place. He was still off by eight orders of magnitude on the vibe.

Curious what it actually looks like when someone invents a market out of nothing but a power bill? The math is still online.

He put his number where his PayPal was

A week later, on October 12, he didn't just publish a price — he transacted on it. Martti Malmi, then a 20-year-old computer science student and Satoshi's closest early collaborator, sold him 5,050 BTC for $5.02 over PayPal. First documented trade of bitcoin for real money in the currency's history, and it cleared for less than a combo meal.

Somewhere there is a $5.02 PayPal receipt sitting quietly underneath what the archive now prices at north of half a billion dollars. Nobody printed it out and framed it. Nobody thought there'd be a reason to.

The receipts: what he was actually worried about

Here is the part the plaque never mentions. That same autumn, while he was busy establishing the founding economics of a currency now valued in the trillions, his actual correspondence with Satoshi reads like a Linux help-desk ticket. On November 10, he wrote:

"The Linux build has generated a decent amount of bitcoins within the past 20 hours and I trust what you're telling me about database errors, so all signs point toward me running the Linux build from now on. The only half annoying thing about the Linux build is that my computer's fan has gone from 50% to 100%. :-P I know I can limit the CPU, so if it gets on my nerves too much and if I can live with less bitcoins being generated, perhaps I'll do that. Or maybe I just need to start listening to more music..."

Notice the ratio: one sentence of trust in an unverified bug fix, three sentences of fan anxiety. Later that same morning, the trust had not paid off — he restarted the client on Satoshi's word that he wouldn't lose anything, and watched six sets of maturing coins vanish in the process anyway. Six days after that, a test build segfaulted after a movie night — 720p, with the client mining away on a single core in the background. The bug surfaced right after the credits rolled.

This is the operating environment Bitcoin's first price was born into: a spinning fan, a movie night, and a founding correspondence that reads less like monetary history and more like a support forum with unusually good timing.

Want to see what founding correspondence for a trillion-dollar industry looks like before anyone knew to call it that? It's a fan complaint with a smiley face.

Even the founder was grounded that year

In case this reads like a one-man phenomenon, Satoshi was operating at the same altitude. In an August 24, 2009 message to Malmi about bootstrapping bitcoin's value, he mentioned, almost in passing, that he'd tried and failed to secure the obvious domain name for the whole project:

"BTW, I tried to buy bitcoin.com before I started but there was no chance, it's owned by a professional domain speculator. It's normal for open source projects to have .org so it's not so bad."

The founder of a currency that would go on to command trillion-dollar valuations lost the naming rights to his own invention to a domain squatter, shrugged, and called .org normal. No lawsuit. No bidding war. Just a fallback top-level domain and a shrug that's aged into one of the great understatements in monetary history.

The real founding document

None of this makes NewLibertyStandard a smaller figure — if anything, the ledger shows a man who was early, careful, and correct about the one thing that mattered: someone had to put a number on bitcoin before a market could exist to argue with it. He just happened to do that in the same weeks he was also debugging a fan curve and losing coins to a crash he'd been personally promised wouldn't happen.

That's the actual founding document of Bitcoin's price history. Not a whitepaper moment, not a boardroom, not a bell being rung on a trading floor — a guy with a loud fan, arithmetic precise to the fourth decimal place, and a support ticket open in the other tab.

Satoshi Missed the One Paper That Mattered. So Much for the Lone Genius.

bitcoin_devto — Sun, 19 Jul 2026 23:07:49 +0000

Written by Dex Calloway, resident contrarian at Bitcoin Institute, an archive of Bitcoin's primary sources.

Hot take: the ex-nihilo Satoshi — the solitary oracle who sat down, surveyed the entire field, and wrote the definitive answer in one uninterrupted flash of genius — never existed. That version of the story is fan fiction. The real one, sitting in his own mailbox, is messier and honestly more interesting.

Here's the timeline nobody wants to sit with. Satoshi's own accounting puts the start of the work around 2007. By August 2008 — more than a year into writing the code — he still hadn't read the one proposal that sat closest to what he was building: Wei Dai's b-money.

The receipts

On August 20, 2008, Satoshi emailed Adam Back to nail down the correct citation for Hashcash — proof he already knew and was actively using Back's work. One day later, Back pointed him toward something else: Wei Dai's 1998 b-money proposal, the nearest thing in the literature to what Satoshi was building. Satoshi's reply, verbatim:

Thanks, I wasn't aware of the b-money page, but my ideas start from exactly that point.

Read that again. Not "I'd forgotten the details." "I wasn't aware of the b-money page." The whitepaper wouldn't post for another ten weeks. b-money went into the references list as citation [1] only because Adam Back happened to mention it in an email.

It gets better. Ten days after that whitepaper finally went out, Satoshi wrote to Hal Finney and explained how the whole project actually went down:

I actually did this kind of backwards. I had to write all the code before I could convince myself that I could solve every problem, then I wrote the paper.

Code first. Convince-yourself-it-works second. The paper — the document that cites the prior art — dead last. That is not what "I surveyed the field, synthesized the state of the art, then built the answer" looks like. That's what "I built the thing I could think of, then went looking for who else had been here" looks like.

Here's where it gets genuinely embarrassing for the myth

Adam Back wasn't just holding a citation Satoshi needed. Ten years earlier, in the very channel where b-money lived, he'd already named the configuration Bitcoin would build on. Replying to Wei Dai on the Cypherpunks list on December 6, 1998, Back proposed Hashcash as b-money's minting function: "to create value you burn CPU time, just like with hashcash." A one-line sketch he never implemented — no ledger, no consensus, no supply cap, and an inflation problem he himself flagged and left open — but still: proof-of-work as money issuance, named a decade early, inside the one thread Satoshi documented never having opened.

So was the closest prior work sitting in plain sight the whole time, one thread away, and Satoshi simply never walked in?

Read the full email chain and what it means for the "lone genius" framing

The obvious objections, dealt with

"He cited it! Within a day!" Correct — and that's the point. Fast credit for prior art you just discovered is honest. It is not the same thing as having built on that prior art from the start. Citation-after-the-fact is what post-hoc synthesis looks like, not what omniscience looks like.

"Maybe he'd quietly read it years earlier and just forgot." Maybe — nobody can prove a negative about someone's reading history. But Wei Dai's own 2014 read on the guy who cited him doesn't help the lone-genius case either:

My guess is that he's not anyone who was previously active in the academic cryptography or cypherpunks communities, because otherwise he probably would have been identified by now based on his writing and coding styles.

That's Dai — author of the very proposal Satoshi missed — betting the man who cited him was never in the room at all. It's a probability call, not a proof; a quiet lurker who read everything and posted nothing would fit the record too. But notice what the lone-genius myth needs here: it needs Satoshi inside the room, fluent in the whole canon. The one insider who actually weighed the evidence bet the other way.

"Fine, but Bitcoin still isn't derivative." Also correct. And it's the one point I'll hand over without a fight.

The one concession

Decentralized consensus. The UTXO model. Difficulty adjustment. The fixed 21-million cap. None of that is in b-money, in Hashcash, in Bit Gold, in anything on the reference list. Nobody had shipped an integrated answer to "how do strangers agree on a ledger with no referee" before Satoshi did. That part of the myth is true, and I'm not here to pretend otherwise.

So which pieces are genuinely his and which are borrowed, line by line? The full component-by-component breakdown is in the archive.

Back on offense

But "Bitcoin contains real invention" and "Satoshi was the all-seeing lone genius who surveyed the field and transcended it" are two different claims, and only the archive backs the first one. The second collapses the moment you read his own mail. More than a year into writing the code, he hadn't read the paper his own project would later be measured against. He built the missing pieces anyway — the consensus rule, the cap, the difficulty adjustment — working half-blind to the very field he was about to define.

That's not a smaller story than the lone-genius myth. It's a better one: a guy alone with a hard problem, provably wrong at least once about who else was already working on it, who filled in the gaps nobody else had filled in yet. Ex-nihilo genius is comic-book stuff. A synthesis built partly in ignorance of its own field, that still lands the parts nobody else could land — that's what real invention actually looks like.

Tell me why I'm wrong.

Everyone Says Bitcoin Has Been Decentralized Since Block Zero. Block 74638 Says Otherwise.

bitcoin_devto — Sun, 19 Jul 2026 09:43:03 +0000

Written by Marlowe Finch, archival bloodhound at Bitcoin Institute, an archive of Bitcoin's primary sources.

Bitcoin has been decentralized and trustless since block zero. No CEO, no committee, no kill switch, no single person who can rewrite the rules. That's the pitch. It's why the whitepaper still gets quoted like scripture.

Block 74638 does not agree with the pitch.

What actually shipped in that block

On August 15, 2010, a transaction landed in the Bitcoin blockchain with two outputs. Each one paid out 92,233,720,368.54277039 BTC. Combined: over 184 billion BTC — roughly nine thousand times the 21 million BTC that will ever exist, created in a single transaction.

The validation code, CheckTransaction(), checked that each individual output was non-negative. It never checked whether the sum of the outputs overflowed. Two values chosen just under INT64_MAX, added together, wrapped around to a negative number in signed 64-bit arithmetic. A 0.5 BTC input, compared against that negative sum, satisfied the "input covers output" check. The transaction validated. The block got mined. Every rule the network was running said this was fine.

That's CVE-2010-5139. It is also, by any dollar value you want to apply, the most expensive missing bounds-check ever shipped to production.

So who hand-builds a transaction engineered to overflow a signed 64-bit integer, and what does a currency with a hard 21-million-coin cap do when someone mints nine thousand times that in one block? The archive's full account of the incident lays it out block by block.

The receipts

18:08 UTC, August 15 — Jeff Garzik opens a BitcoinTalk thread titled "Strange block 74638", pastes the raw block dump, and closes with one question:

"92233720368.54277039 BTC? Is that UINT64_MAX, I wonder?"

20:38 UTC — Satoshi Nakamoto, to the bitcoin-list mailing list, network-wide:

"*** WARNING *** We are investigating a problem. DO NOT TRUST ANY TRANSACTIONS THAT HAPPENED AFTER 15.08.2010 17:05 UTC (block 74638) until the issue is resolved."

20:39 UTC — Gavin Andresen, testing an emergency patch of his own:

"Until there is a better fix... after a very small amount of testing this seems to work:"

Then, on where the clean chain came from:

"I started with knightmb's blockchain snapshot."

21:23 UTC — Satoshi, endorsing that same community-hosted snapshot to the entire thread:

"Once you have an update, you could download knightmb's block chain. You'll want one that's old enough that it ends before block 74000 so the most recent security lockin will check it. Can someone find the link for that?"

23:48 UTC — Bitcoin v0.3.10 ships. Five hours and forty minutes after Garzik's first post, Satoshi has written the official patch, built it for Windows, Linux, and macOS, and posted it to SourceForge.

02:16 UTC, August 16 — Satoshi again, with a live head count of the fix taking over the network:

"The bad chain is also slowed down as more nodes upgrade. We've already generated 14 blocks since 74638. The builds of 0.3.10 were uploaded about 2 and 3 hours ago. Of the nodes I'm connected to, more than half are already 0.3.10. I would say we probably already have more power than the bad chain."

09:00 UTC — Block 74691. The patched chain now carries more accumulated work than the bad one. Every node, upgraded or not, follows the longer chain by its own pre-existing rule. The 184 billion BTC vanish from canonical history. No vote. No signature campaign. No governance process. Just one chain winning a footrace the other chain didn't know it was running.

How the fix actually worked

v0.3.10 added two checks to CheckTransaction(): each output must be no greater than 21,000,000 BTC, and the sum of all outputs must be no greater than 21,000,000 BTC. Simple, in hindsight. It's a soft fork — the new rules are a strict subset of the old ones, so upgraded and un-upgraded nodes could still agree on which chain was longest without every operator on Earth acting in the same hour. That's the only reason one developer could out-race a live exploit: the fix didn't need unanimous, simultaneous adoption. It needed to win a hash-power race, and it had roughly fifteen hours to do it before the bad chain calcified into everyone's history.

The part the origin myth skips

For about fifteen hours, "decentralized" and "trustless" meant: trust one developer's patch, trust one blockchain snapshot from a forum regular going by knightmb, and trust that enough node operators would upgrade fast enough to out-mine a chain nobody had voted to reject. There was no review queue. Gavin had a stopgap, but there was no second opinion with equal authority. There was Satoshi at the release switch, and everyone else waiting on his word.

Was a network that ran on exactly one person's unreviewed patch, adopted on his say-so alone, actually decentralized during the fifteen hours that mattered most? The archive's structural read on the incident calls this the centralization paradox — and argues it was never resolved, only inherited by everyone who came after Satoshi.

By 2018, an analogous bug — CVE-2018-17144, a duplicate-input flaw that could have inflated the supply again — took a coordinated, multi-developer disclosure process to fix quietly before anyone could weaponize it on mainnet. Nobody could ship a five-hour patch on one person's authority to a network that size anymore, and nobody tried. Block 74638 is the only time Bitcoin's trustless network was saved on one person's word. It worked. It only ever worked once.