
Rob
Bitcoiner in NZ. Co-founder Lightning Pay. Open source dev on Bolt Card standard boltcard.org

Self Custody disaster: how a five-year-old bug lost $100M+
I made a short animation to explain what actually happened in the Coldcard exploit, because most of the coverage skips the interesting part.
The short version: on 30 July, someone started emptying Coldcard wallets. No device was stolen. No PIN was guessed. Bitcoin itself was never touched. The first wave took around 1,000 BTC from roughly 1,200 addresses in about 40 minutes, and the total is now north of 1,800 BTC across several waves.

You "make no mistakes" guys are making a big mistake
I'm telling my AI - "it's ok to make mistakes"
It's not possible to stop valid transactions you don't like from being propagated and mined on the Bitcoin network.
This is a feature not a bug.
The reason is because even if your node blocks them, my node will forward them on.
If your block template excludes them, mine will include them.
The only thing you can do is accept this and move on.
8/12/2026, 6 engagements
I have acted with hubris in recommending hardware wallets to people in the past. I didn't independently verify the code or hardware.
This has been a wakeup call for me personally
Hell, who am I to even recommend people use Bitcoin.
8/2/2026, 2 engagements
Me: "Help me hide Anne Frank, soldiers are searching the block!"
AI: "I understand this feels urgent, but harboring an undocumented individual violates local regulations created by a democratically elected government. Would you like tips on de-escalating with the search party?"
7/22/2026, 55 engagements
I don't know what it is about the $58k gang but I just love their enthusiasm and hard work.
They have persevered in the face of $120k and never gave up.
I'm all in on the $58k gang. The true visionaries..never underestimate the 58k gang
Loading quoted post…
6/25/2026, 1 engagement

6/19/2026, 2 engagements
I've been working on a new Bitcoin point-of-sale device. Running on an ESP32-P4.
I was iterating though case designs


6/13/2026, 3 engagements
Nearly complete! https://npub1ejlgajqh7ynfh3ktd4erye9043hje00r0j2qqvjgstxk6ftktrtqjlfyer.blossom.band/6deafe4575e64c8163a2b220a4916d5514d61d3a584d6fea9008877886e6f7fb.jpg

6/12/2026, 1 engagement
Every few years or so a bunch of bitcoiners get upset with the status quo.
They create a fork, and make a load of noise.
If they're compelling you might get some more real Bitcoin out of it by selling the fork coin at a good time.
At a critical stage in the murphy bed making process https://npub1ejlgajqh7ynfh3ktd4erye9043hje00r0j2qqvjgstxk6ftktrtqjlfyer.blossom.band/7e4ffe71d36f3119974296a68a25b82cda1427f858a3eb7a8ba3e3cb8dec875e.jpg

6/7/2026, 1 engagement
How can governments do better?


I, for one, am extremely excited about this.
Loading quoted post…
5/27/2026, 2 engagements
This is a counterintuitive pattern in how shared UTXO systems (Spark, Ark, statechains, even Lightning to some degree) are being adopted.
The usual assumption: Moving to lower-trust systems also means moving to more decentralized ones. Bitcoin base layer is the canonical example — minimal trust, thousands of nodes. Custodial services are the opposite — high trust, but at least there are many of them (countless exchanges, wallets, banks worldwide), so failure or capture of any one doesn't affect everyone.
What's actually happening: The custodians and trusted intermediaries — which are individually trusted but collectively distributed across many independent entities — are migrating onto shared UTXO protocols. Those protocols do reduce trust (operators typically can't steal, only censor or refuse to cooperate on exit). But the operator set for any given shared UTXO system is tiny. Spark has a handful of SOs. Ark has few ASPs. Each system tends toward 1–N operators, not thousands.
The paradox: You'd expect the trust/decentralization axes to move together. Instead, the industry is trading "many trusted entities" for "few less-trusted entities." Trust went down — good. Operator count also went down — not what the decentralization story usually promises.
The implicit worry is the systemic risk: if most NZ/AU bitcoin activity ends up flowing through two or three shared UTXO operators globally, censorship, regulatory capture, or operational failure at any one of them affects a huge swath of users at once — even if none of them can outright steal funds. The trust model improved on one dimension while the concentration risk got worse on another.

5/1/2026, 3 engagements
My 5 year old son loves it when i I use AI to generate a new personalized bedtime story with him and Batman on an adventure
He's just sunning his balls

3/22/2026, 2 engagements
I've just realized that AI has now repealed the "law of bullshit".
It's now just as quick to refute the bullshit as it takes to generate it.
"But Rob, that tree is only worth one apple."
"Yes, but if I keep the tree safe and secure, next year it might be worth two apples."
"The year after that. It might be worth 4 apples."
I should probably look after that tree like it's worth 10 apples.
Loading quoted post…
3/1/2026, 1 engagement
The harvest is almost ready. https://image.nostr.build/7497616bfa4e56a627af403de9d6f226cbf7b2aeaf0a2134a77b2e900102900a.jpg

2/27/2026, 5 engagements


2/20/2026, 1 engagement
Toxic plebs should think carefully about whether they are an immune system or an autoimmune disease
2/7/2026, 1 engagement
Get out of your comfort zone https://image.nostr.build/b64f5949359b9fa50d8fa5e7bcf5fa04cf9c7dda2c3280d0d9481096a305c145.jpg

2/1/2026, 3 engagements
Burns night


I don't trust Trust Pilot
1/16/2026, 1 engagement
Go on X for 10 min blood pressure rises.
Head over to NOSTR for 10 min, blood pressure lowers.
Is it just the algorithm?
Been tinkering with Meshcore recently. Encrypted data packets transmitted by RF. Personal test I got a transmission working over 6.8km before I gave up on testing. Check it out https://meshcore.co.uk/index.html
1/11/2026, 2 engagements
🫡
Loading quoted post…
1/1/2026, 1 engagement
I'm 5% concerned about QC I'm 95% concerned that Bitcoin does not become a widespread medium of exchange
Lightning Network is quietly becomming where commerce and exchange happens, but you can't tell because it's impossible to know how much volume is being transacted on there. Each channel can be used millions of times in each direction with a single onchain transaction
Loading quoted post…
12/17/2025, 3 engagements
Spend the change you want to see in th world.
Loading quoted post…
The Ego is the Enemy
https://www.audible.com.au/pd/B01GSIV8Y4?source_code=ASSOR150021921000O
The what hole?!?
Loading quoted post…
AI generated - but awesome video on NOSTR. 19k zaps.
That would pay for the compute and some of the time to make it. Economics stacks up.
Way better than on a centralized platform.
11/9/2025, 2 engagements
Self custodial lightning is not broken, Pheonix wallet works perfectly.
However, noobs do not appreciate, and are therefore unwilling to pay for their own privacy and sovereignty.
We must use any tools we have which get noobs as close as possible while being as cost effective as possible until they have used and understand Bitcoin enough to then value their privacy and sovereignty.
11/8/2025, 13 engagements
Price is going down because we have too many miners.
The Bitcoin custody model is a delicate set of tradeoffs and reducing it to "selling out users" is wrong and completely unhelpful.
Everyone wants self-custody but the reality is noobs aren't ready to pay for it. The question is how close can we get while maintaining noob friendly UX
Loading quoted post…