Mat Balez

This morning, my 7-year-old is off to the dentist for a filling. My eldest is staying home with a nasty cough. Meanwhile, my daughters are dressed as bumble bees and unicorn fairies. And it’s day #1 at @spiralbtc. It might not seem like it, but all these things are the dream.

thx for sharing the piece. i think escrow and delegated spending with control are interesting examples, questions in my mind are: 1) how important are these for high volume tiny transactions? 2) how much of such needs can be adequately (or better?) be handled outside the protocol, say via wallet layer, app layer, or some competing 3rd party services acting in a permissionless network step in to fill the need (for instance for escrow or insurance or other needs where it makes economic sense for other actors to get involved) 3) similarly, will reputation systems be developed that obviate the need for in-protocol trust-minimization beyond settlement?

@matbalez I very much doubt it will be this simple. Even with simple payments, conditional terms may apply and be managed by agents by ways of escrows and other systems. And payments are just scratching the surface of what agents will do with money... x.com/ArkLabsHQ/stat…

Arkade is the single best version of Bitcoin for agentic commerce. Our intent system is built-in to support the AI paradigm and allows composable, programmable actions that aren't covered by other scaling protocols. Of course, it taps into Lightning for interoperability needs.

@bergealex4 i want to think more about programmability and how important that is aren’t the vast majority of agentic payments just: - i want thing that costs money - pay for thing - get thing what is the big programmability unlock?

@matbalez if the agent loses money because of channel state issues, that's a real cost. arkade is self-custodial, cashu is not so there's that. and I know there are cool stuff you can engineer to program cashu payments but that doesn't seem to be getting standardized anytime soon.

@bergealex4 what is your assessment of arkade approach vs cashu? fwiw I tend to think agent “preferences” will need to be expressed in things that can be calculated: real costs, risks, etc it’s not clear to me that an agent will develop a preference based on the virtue of “statelessness”

@matbalez 1. Programmable 2. Easier spending delegation 3. Stateless 4. Opportunity cost of liquidity management Channels are suitable for high-frequency bi-directional flows. A UTXO can ride Lightning's network at any time if needed but becomes less versatile locked into a channel.

@bergealex4 what is your steelman of bearer money over liquidity channel for agents?

@matbalez I'm skeptical that agents will prefer liquidity channels to bearer money. That said, Lightning has significant traction as a standard payment format within Bitcoin. Too early to tell whether or not it's the right approach but it's a good neutral interoperability standard.

@bergealex4 thx will try this stuff out what are your thoughts on L402 as a way for agents to transact?

There is! We currently use intents on Arkade using the BIP322 framework to provide proofs-of-ownership and manage/delegate participation in batches. This can be extended down the line to more programmatic flows using contract templates. Bearer assets and UTXOs are simply much better suited to agents than the alternatives. clw.cash also github.com/arkade-os/skil…

new metric just dropped

It’s time for Bitcoiners everywhere to build. Every player in the payments space is starting from zero and we have a real shot at actually taking some market share from huge centralized intermediaries. But if we want a shot we all have to pitch in, and build momentum fast.

Can’t make the NY event this time around, but I went in SF and it was 🔥 Nothing beats a room full of peers vibe coding together. Electric energy. Also @HBerkoe @moneyball @matbalez IRL ✨ luma.com/lk606302

probably the quirkiest thing i've ever done with bitcoin: i sent a potato to google HQ in mountain view! spuds.lol clean build by Flint member @austinvach stay weird friends

@BitcoinCoderBob @SahilC0 @tmknsm @moneyball sats ship has sailed

@SahilC0 @matbalez @tmknsm @moneyball 1,000,000,000 sats 2,500,000 sats 107 sats Don't need to abbreviate, don't have to repurpose symbols that already have established meaning (doing so is actually more confusing)

say we wanted to adopt bip 177, but the product has a mix of large amounts (10 BTC loan), medium amounts (spending balance ₿2.5M), and small amounts (₿ 107 fee) how should this work? @matbalez @tmknsm @moneyball im thinking little things like alignment between units in a list

@programmer has a fix for the wallet startup issue been shipped yet @programmer? my wallet still seems to have this problem.

@programmer thanks for the response! was able to work around the issue with xvfb. will stay tuned for actual fix.

Agent's need to: - Send - Receive - Pay - Trade Now they can. Introducing awal, an agentic wallet purpose built to AI. Try it for yourself: npx awal@latest --help Give it to your agent, they can figure it out: npx skills add coinbase/agentic-wallet-skills 🧵

@StephenDeLorme @SahilC0 @tmknsm @moneyball yes in visual form: x.com/matbalez/statu…

I read your original description, but just because there is a mix of large and small numbers, it doesn't follow to me that we need to change the units just for the size of the number. Loan Balance ₿ 1,000,000,000 Spending Balance ₿ 2,500,000 Today's Fees ₿ 107 That looks fine to me. If you run out of room in the UI, then just truncate according to a localized standard. In the US that's: Loan Balance ₿ 1B Spending Balance ₿ 2.5M Today's Fees ₿ 107 Even better, we can absolutely avoid confusion just by putting some local fiat valuations in there. Loan Balance ₿ 1B, ~$655K Spending Balance ₿ 2.5M, ~$1.6K Today's Fees ₿ 107, ~$0.07 For customers who really want to see BTC, that's fine. I think currency and localization selection is a pretty common design pattern. I switch all my devices to 24 hour time. Just have a switcher to show things in BTC.

@SahilC0 @StephenDeLorme @tmknsm @moneyball

@StephenDeLorme @matbalez @tmknsm @moneyball imagine a loan product where there are large amounts but also small amounts like fees or payments

@janrothen @SahilC0 @tmknsm @moneyball ₿2.5M = 0.025BTC = 2.5M sats (deprecated)

@SahilC0 @matbalez @tmknsm @moneyball What does ₿2.5M stand for (amount in sats)?

@nickslaney @nvk @JustinOHallo @moneydevkit @StephenDeLorme i've had a lot of fund building apps with MDK and latest obsession is using their new L402 stuff to build agent experiments

@nvk @JustinOHallo @moneydevkit @StephenDeLorme @matbalez Hey happy to help here, I'll dm

Dear Lighting friends. Replace Stripe without KYC and I'm your customer at 2% fees. Even better, make it API drop in replacement. That's your competition. With KYC, credit cards are infinite more customers and less complexity. Baselayer is already a winner.

@bitbrunchpod This is something pretty meta we have been getting on about lately: That natural language is now the language of the machines: words themselves And those who are the most powerful with words, i.e. philosophers will inherent the world.

this has my attention agents can buy domains entirely on their own now no manual clicking around no account or log in no credit cards no identity a glimpse at the future of commerce but it’s today

if open-source AI doesn’t build real agent-ready payment rails now, future AI commerce will be controlled by big tech payment stacks instead of the open internet. 🚀🧠💳 love @TheBlueMatt rage on this.