6 August 2026 — Amir Sadjady & Eric Pace, with Ray Amir opened with the Coldcard hack. Not a Dtravel topic, and the best call of the hour. For anyone who missed it. On 29 July, thousands of C
6 August 2026 — Amir Sadjady & Eric Pace, with Ray
Amir opened with the Coldcard hack. Not a Dtravel topic, and the best call of the hour.
For anyone who missed it. On 29 July, thousands of Coldcard hardware wallets were drained without anyone touching them physically. Amir's figures, hedged as he gave them: somewhere over $100 million, maybe 5,000 wallets. The cause was the random number generator used to create seed phrases. That's where it went wrong.
Worth understanding who this hit. Coldcard was regarded as one of the most secure wallets on the market and it only worked with Bitcoin. So its users were maximalists, dollar-cost-averaging for years, sticking the device in a safe or a deposit box and never touching it. This was pension money. Generational-wealth money. The people who did everything the careful way.
Then a quote went round, from a man named Goodman. Society only works because of trust. Cross a bridge, you're trusting the architects built it properly. Drive a car, you're trusting the manufacturer. He'd done nothing wrong and had his livelihood shattered anyway.
Amir wasn't bringing it up to kick the manufacturer. More the other way round.
Eric's answer to that was the sharpest thing he said all hour. The promise of crypto was that you could stop trusting people and trust math instead. What broke wasn't the math. It was everything wrapped around it — the firmware, the build process, the auditors, the people who wrote the code, the entropy source.
And then the line worth keeping: instead of a bank you could sue, it's now a codebase created by people.
Trust the math, not the person. Which, fine, yes. Except the math is running on a device somebody built. You can push trust as far down the stack as you like and it's still sitting there at the bottom. Just harder to see. Much easier to stop mentioning.
Everyone stopped mentioning it. Crypto moved trust. It didn't remove it. The question is where you'd want to put it, and Eric's framing of that was that these are micro-decisions people make constantly, mostly without noticing — the bridge, the car, the wallet.
Then Amir went back to his mother's Airbnb listings. Four of them, delisted overnight, ten years ago. She'd done nothing. No reason given. This was before Airbnb built its automated appeal process, so there was nothing to appeal to. He wrote emails. Nobody answered. She can still log in and see the listing sitting there.
He's told this one a lot. It lands harder each time rather than softer.
And it's the same shape as the wallet, which is the whole point of putting them next to each other. She opened the business trusting it was a safe place to operate. That trust was the deal. It broke, and there was no recourse in the design.
The lock-in follows from that. Reviews aren't portable. Years of work building a rating and you can't take any of it with you, which leaves Airbnb as the adjudicator of whether your business gets to exist. Amir's read: nobody has solved this, and Airbnb isn't trying to.
Eric added the part people miss. It isn't that hosts love Airbnb. It's that if you want your first customers, there are maybe two or three places to start, and realistically one. Travellers face the same short list. So the trust gets placed subconsciously, because there's nothing much to choose between. Dtravel's actual challenge, in his words, is building an alternative and then giving people enough confidence to try it.
Amir's point, and he's been circling it for weeks: people aren't going to Google Flights any more. They're going to ChatGPT, Gemini, Claude, and writing real prompts — split legs, cheaper routings, the lot. Small group so far, and ahead of everyone else because of it. He doesn't think accommodation is far behind.
Eric called it a race, which is the right word. The funnel has been fixed for twenty years. Google, then an OTA, then a booking. That's coming apart, and the agent becomes the top of the funnel.
His questions were the good ones. How does anyone know an agent's results are trustworthy? How do you know the link it hands you wasn't planted by a fraudster? How do agents scrape sources in the first place, and what's the framework for any of it?
Nobody's answered that yet. It's the piece of the thesis that doesn't need talking up — the rest takes argument, that one is just true.
Reputation built from verified actions, in the network, readable by an agent deciding who's real. Amir admitted he didn't get it at first and came round to it after doing his own research, which is a more useful thing to hear than enthusiasm.
Eric was careful, and this is the bit that matters. There is still a trust system. It doesn't vanish. What changes is who runs it: instead of one company holding the keys, employing the trust team, making calls it won't explain because it's protecting its legal position, you get community governance. It might keep some of the same elements. The difference is that anyone inside the system can see it and take part.
The whitepaper is where this gets fleshed out. Eric said soon. The commitment made on the call was autumn, and it should be read as a deadline rather than a hope. Everything else described here depends on it, and a rebirth that can't produce the document explaining itself isn't a rebirth.
Amir has been made Director of the Dtravel Foundation, and he made a point of saying he's never worked for Dtravel. He came from the community.
The argument for the DAO ran through disputes, which is where it gets concrete. Right now a problem with Airbnb sends you to a resolution bot. If you push, you might get a human, and that human is bound by guidelines they can't break — so you never actually get human judgement, just a person reciting policy. Reputation belongs to the company, and whatever the company says goes.
If reputation is going to be the thing people build their businesses on, disputes need a neutral way to resolve and both sides need standing. That's the case for participants co-owning the layer rather than renting it. Dtravel is on course for the transition. It comes after the whitepaper, and unlike the whitepaper it doesn't have a date on it yet.
Amir put the first one to Eric directly: if a trustless hardware wallet still requires trusting the people who built it, isn't Dtravel just asking people to trust it in the same way, with extra steps?
Eric's short answer was yes. Trust is essential to everything in society, the travel economy already runs on trust systems, and the question is only whether a better one can be built and whether that's a better place to put it. It was a brief answer. Honest, and briefer than the question deserved.
The second was harder. Goodman did everything right — self-custody, a reputable hardware wallet — and still got hit. What stops that happening to someone's reputation on Dtravel?
Eric didn't dress it up. Nothing is unhackable and he wasn't going to make false promises. What you can do is design differently, and that means eliminating single points of failure. Centralised design carries inherent risk, and he thinks that risk climbs as AI capability accelerates. His view is that centralised is already the risky option today.
Which is a real answer and still isn't an answer to the question that was asked. Fewer single points of failure is an honest sentence. It is not the same sentence as you're safe. Better to keep saying the first one badly than the second one smoothly.
Ray came up to speak. He was a big Dtravel investor at the end of 2021, held a lot of the token, and everyone can see what the chart has done since.
His argument, and it's his rather than the team's, is worth quoting properly. Young crypto investors mostly aren't in it for utility or governance. They buy because of momentum, because of a recommendation, because of YouTube influencers and social marketing — and the utility follows later, if it follows. He also said he's a host and an Airbnb user himself and doesn't much care about the governance side. He participates for the rental.
Then the line of the call: for him, the chart dynamics are the trust issue.
Amir agreed with all of it, and said so as a crypto person rather than as a spokesman.
This was the most candid stretch of the hour and it hasn't been laid out this plainly before.
TRVL was designed as a rewards token, modelled on what works for Travala. The decentralised booking layer got built and it works — bookings come in every day, visible in the team's Telegram. But the network wasn't growing fast enough for a rewards token to mean anything, the alt season everyone was waiting for never arrived, and people dumped.
Last year there was a real decision point, because the people running it wanted to walk away and do something else. Eric didn't. He took over as CEO on the basis that the core idea was still worth something. Amir's view is that Dtravel has only now found the identity it needed from the start.
Amir also asked the honest version of the question out loud: the network is built, it works, so why aren't people coming? He didn't pretend to have found that easy to answer.
This is the part that answers Ray, and it's new.
Dtravel takes a share of the fee revenue an action generates, buys TRVL with it on the open market, and locks it into that person's standing permanently. Those tokens leave circulation and don't come back.
What that's worth isn't something to characterise here, and it would be a mistake to try. The mechanism is real and at current revenue it is small. Both are true at once, and the whitepaper puts a number on it rather than an adjective.
Amir was clear that this needs the whitepaper before it will land, because until people can see where value comes from they won't understand the why behind the token. Which is the whole reason that document has a deadline on it rather than an intention. He also made a broader claim: Dtravel is a proof of concept. If a portable reputation layer works here, it exports to anywhere else AI needs to know whether a business or a person is real and has a record behind them.
The Space moves to every other week. Next week is off. It also shifts about three hours later, on the basis that the analytics say that catches more people.
Separately, Dtravel content is coming off the Sunday and Tuesday livestreams and getting its own video, made when there's something to report. A weekly slot makes you produce updates because it's Thursday rather than because anything happened, and everybody can tell the difference. They just don't say so.
Amir's own closing, near enough word for word:
Whether it's a hardware wallet or a platform holding your reviews hostage, the lesson is the same. Trust doesn't go away, it just moves. The only real question is where it lands and whether you can check it.
