Citadel Dispatch

UTXO The Webmaster joins to discuss his new android nostr client Wisp, spark wallet integration, encrypted nsec seed backups, and his controversial "Send Money" normie mode that denominates zaps in dollars. We get into follower count philosophy, what actually counts as a user on an open protocol, and why half of bitcoin twitter still hates nostr. Then we go deep on AI: his local rig running Qwen 3.6, on device spam filtering with nspam, whether Claude is subsidizing us into oblivion, and what happens to big tech jobs when a unicorn only needs three employees.

UTXO on Nostr: https://primal.net/utxo
Wisp: https://wisp.mobile/

EPISODE: 200
BLOCK: 946079
PRICE: 1320 sats per dollar

(00:02:05) Wisp origin story
(00:02:39) Why Wisp? Stability, UX focus, and the outbox model
(00:05:30) Availability: Android, beta rollout, and Zap Store install
(00:06:14) Onboarding with Spark, Zaps, and wallet backups via nsec
(00:09:02) Custody, risk, and privacy tradeoffs for Nostr Zaps
(00:11:27) Wisp’s "send money" UX and fiat denomination debate
(00:16:39) Do small zaps feel insulting? Behavioral effects of denominating in fiat
(00:17:06) Follower counts: definitions, reputation, and network-relative views
(00:23:07) Bots, fake metrics, and survivorship bias on open networks
(00:31:23) Keys, compromises, and practical key rotation culture
(00:38:05) AI tools in Wisp development and local vs cloud models
(00:45:49) nspam: on-device reply filtering without killing good bots
(00:47:53) Bots with their own feeds, rate limits, and unstoppable relays
(00:53:08) Growth via creators: streaming, ZapStream, and multi-platform outreach
(00:58:58) Mirroring vs authentic posting: does it feel stale or disrespectful?
(01:02:11) Businesses living on multiple platforms and Bitcoin payments
(01:03:04) Daily AI workflow: Claude, APIs, local models, and cost control
(01:06:44) Privacy, local hardware as luxury, and pay-per-query services
(01:11:11) Five-year AI outlook: limits, jobs, bubbles, and lean megacorps
(01:21:26) Closing thoughts and next steps: try Wisp, share feedback

more info on the show: https://citadeldispatch.com
learn more about me: https://odell.xyz
monitor the situation: https://citadelwire.com

Craig Raw, creator of Sparrow Wallet, joins to discuss silent payments, a new bitcoin address system that eliminates address reuse, removes the gap limit, and aligns privacy with convenience. Craig walks us through the history of bitcoin address derivation from single key to hd wallets to bip 47, then explains how silent payments optimizes everything except scanning cost and how his new server implementation, Frigate, uses gpu acceleration to mitigate that. We discuss the path to adoption including hardware wallet support, public server infrastructure, bip 353 human readable addresses, and the overall vision of upgrading from hd wallets to sp wallets.

Craig on Nostr: https://primal.net/craigraw
Craig on X: https://x.com/craigraw
Sparrow Wallet: https://sparrowwallet.com
Frigate Repo: https://github.com/sparrowwallet/frigate

EPISODE: 199
BLOCK: 944916
PRICE: 1384 sats per dollar

(00:03:09) Craig Raw of Sparrow Wallet
(00:03:27) Silent Payments: what they are and why they matter
(00:06:01) From single keys to HD wallets: history and limits
(00:11:41) Address reuse in the wild and UX realities
(00:11:50) BIP47 review: pros, cons, and hardware wallet hurdles
(00:15:18) Enter Silent Payments: design tradeoffs and hardware support
(00:19:01) Key benefits: static codes, enforced freshness, no gap limit
(00:21:02) The scanning-cost problem and early client approaches
(00:25:27) Server-side strategy: database tweaks and GPUs
(00:29:15) Why public servers matter and performance breakthroughs
(00:33:37) Frigate with Electrum backends: deployment paths
(00:37:20) Risks with public servers and practical mitigations
(00:43:10) Uncle Jim model and GPU-ready home servers
(00:46:21) GPU backends: CUDA, OpenCL, Metal and real-world nodes
(00:47:34) Running everything on a laptop and pruning considerations
(00:49:11) Human-readable addresses: DNSSEC and BIP353
(00:55:14) What’s needed next: hardware, node vendors, and runners
(00:58:13) PSBT details, DLEQ proofs, and multisig caveats
(01:02:13) Timeline to usable SP wallets and public servers
(01:06:10) Reframing SP as UX: contacts and everyday payments
(01:07:22) Ecosystem fit: who could ship this first
(01:08:31) Wrapping up: calls to action and outlook
(01:10:03) Closing notes: upcoming guests and events

more info on the show: https://citadeldispatch.com
learn more about me: https://odell.xyz
monitor the situation: https://citadelwire.com

Justin, a prolific contributor to the Fedimint open source project, returns for a six month update. Fedimint is an open protocol providing easy to use, private, programmable, and offline bitcoin payments using bitcoin powered federated chaumian ecash.

Justin on Nostr: https://primal.net/p/nprofile1qqspg8fq209jj56663d2n6r9ehkyjffy7rkqqejfdwvtwzva426avkqxtxxuv
Fedimint Website: https://fedimint.org/
Fedimint on X: https://x.com/fedimint
The Ecash App: https://ecash.love/
Fedimint Observer: https://observer.fedimint.org/
Bitcoin Mints: https://bitcoinmints.com/
Iroh: https://www.iroh.computer/

EPISODE: 198
BLOCK: 944073
PRICE: 1466 sats per dollar

(02:06) Justin on Fedimint updates since last visit
(03:20) Ecash App vision as a Fedimint reference client
(04:18) Wallet features: on-chain, lightning, ecash, and nostr integrations
(06:01) Fedimint 101: federations, guardians, and multisig trust model
(07:55) Uptime vs. rug risk and Byzantine fault tolerance in practice
(09:18) Making guardianship easier and raising operational reliability
(10:14) Ecash App status, platforms, backups via nostr, and seed UX
(13:16) Mint/federation selection challenges and web-of-trust ideas
(15:39) Observability tools and on-chain vs. Lightning differences
(16:20) Running a Guardian on Start9: setup and backups
(19:39) Networking with Iroh: DNS removal, privacy, and Tor/VPN plans
(23:14) Lightning gateways: roles, trust, liquidity, and multi-federation ops
(27:59) Gateway UX: multiple gateways, auto-switching, and agents help
(29:01) Gateway pairing and funding flows for Start9 deployments
(32:24) Guardians on Android phones: why, how, and trade-offs
(37:30) Blockchain backends: Bitcoin Core vs. Esplora defaults
(39:30) Mobile data, heat, and practical considerations
(39:34) Agentic payments and why eCash fits well for agents
(43:40) Local communities, AI models, and community services vision
(46:06) Real-world adoption, roadmap, modules, and BOLT12 plans
(48:50) BOLT12 receive-side challenges and trust model nuances
(50:26) Pragmatic trust, permissioned gateways, and next steps
(50:37) How listeners can help and contact info
(51:18) Start9 v0.4.0 update chatter and flashing war stories
(53:01) Closing thoughts, progress praise, and sign-off

more info on the show: https://citadeldispatch.com
learn more about me: https://odell.xyz
monitor the situation: https://citadelwire.com

Matt Ahlborg, founder of PPQ.ai, rejoins the show for an update on the rapidly evolving AI landscape. PayPerQ is a bitcoin enabled ai platform that enables users to easily use all of the top ai tools without an account. Users pay per use with bitcoin and can switch the models they use on the fly without needing to provide an email address, phone number, or billing address. We discuss the rise of lean teams supercharged by AI tools, the subscription vs. pay-per-token model debate, and why massive subsidies from companies like Anthropic and OpenAI likely will not last. Ahlborg breaks down PPQ's "AutoClaw" smart routing feature that blends cheap and expensive models to cut costs, the addition of secure enclave models for privacy conscious users, and how OpenClaw's explosion drove a 400% revenue increase for PPQ.

PayPerQ: https://ppq.ai/
PayPerQ on Nostr: https://primal.net/p/nprofile1qqsdy27dk8f9qk7qvrm94pkdtus9xtk970jpcp4w48k6cw0khfm06mss64u96
PayPerQ on X: https://x.com/PPQdotAI
Matt on X: https://x.com/MattAhlborg

EPISODE: 197
BLOCK: 942174
PRICE: 1412 sats per dollar

(00:02:57) Matt Ahlborg of PPQ.ai and the fast pace of AI
(00:04:48) Early‑AI "Wild West": workflows, tiny teams, and hiring realities
(00:07:58) Who benefits most from AI? Devs, non‑devs, and the humility to learn
(00:13:00) Two ways to use AI: locked‑in subscriptions vs pay‑per‑token sovereignty
(00:17:46) Business model nuance: subsidies, vendor lock‑in, and PPQ margins
(00:21:00) Open models improve but show limits under real workloads
(00:23:29) AutoClaw smart routing: mixing cheap and premium models
(00:27:12) Routing tradeoffs: cost, competence, latency, and "quarterback" models
(00:31:13) Secure enclaves and privacy: running models in TEEs
(00:38:00) OpenClaw agents: promise, bugs, and the personal AI assistant future
(00:41:22) Building a personalized AI newswire with Nostr and RSS
(00:51:02) Payments debate: Bitcoin first vs accepting everything
(00:58:03) Comparing PPQ and Venice: tokens, privacy claims, and incentives
(01:02:10) Usage data: what users pay with and which models they choose
(01:08:16) Runaway costs and safeguards: spending limits and lessons
(01:08:40) Agentic payments and L402: where Lightning fits vs x402 vs MPP
(01:15:10) Closing thoughts and what’s next for PPQ.ai

more info on the show: https://citadeldispatch.com
learn more about me: https://odell.xyz
monitor the situation: https://citadelwire.com

Evgeny is the founder of SimpleX Chat, a private and secure comms protocol that has a radically different approach to the concept of user identity. We discuss how SimpleX's unique transport network assigns addresses to connections instead of endpoints, why MLS is flawed, the upcoming scalable channels feature to compete with Telegram, and how the network plans to sustain itself through a model where large channels fund infrastructure. No phone numbers. Private and secure. Open and scalable.

Personal blog: https://www.poberezkin.com
Official website: https://simplex.chat
SimpleX on Nostr: https://primal.net/simplex
SimpleX on X: https://x.com/SimpleXChat

EPISODE: 196
BLOCK: 941454
PRICE: 1432 sats per dollar

(00:02:56) Introducing SimpleX and why Signals model falls short
(00:04:48) What is SimpleX? Sovereignty and trustless design principles
(00:09:21) Privacy as prerequisite for speech and society
(00:13:04) From messenger to scalable channels and Telegram comparisons
(00:17:27) Content privacy vs participation privacy in large groups
(00:23:30) Removing identity
(00:24:32) Transport layer innovation: addressing connections, not endpoints
(00:29:09) SimpleX Chat as first app and platform on the network
(00:30:25) Agents, AI, and commerce inside messaging
(00:32:43) Routers: resource needs and the trust model
(00:36:14) Operator diversity and Tor comparisons
(00:40:15) Packet level anonymity vs persistent circuits
(00:41:39) Discovery and first contact: addresses, reply paths, UX
(00:43:09) Groups at scale, MLS critique, and Signals approach
(00:48:00) SimpleX groups today and upcoming channel relays
(00:52:30) Verifiability, signed actions, and deniability tradeoffs
(01:01:02) Authenticity for public speech in a deepfake era
(01:02:01) Incentivizing infrastructure: beyond hobby servers
(01:08:10) Why premium app models fail; web monetization analogy
(01:11:00) Channels as websites: who pays and why
(01:14:34) For profit vs nonprofit: incentives, governance, and scale
(01:21:16) Consortium governance and resisting capture
(01:27:41) Lessons from the web: speed, innovation, and standards
(01:33:06) Privacy tech adoption realities and movement unity
(01:34:36) Monetization mechanics: registries, naming, and smart contracts
(01:39:54) Programmatic revenue sharing and prepaid credits
(01:52:18) Choosing chains and assets: centralization vs volatility
(01:55:09) Prototype first, prove market fit, then harden design
(01:59:00) Motivation: restoring private communication at scale
(02:00:12) Next steps: consortium, crowdfunding, and closing

more info on the show: https://citadeldispatch.com
learn more about me: https://odell.xyz
monitor the situation: https://citadelwire.com