18: Rewriting TENEX w/ Pablo

Article header image for 18: Rewriting TENEX w/ Pablo

Pablo rewrote TENEX about a hundred times. Multi-agent coordination on Nostr, vibe-coded NDK ports, and why you're basically yelling at someone who is lying to you.

Author: NoSolutions
Published:
Format: Markdown (kind 30023)
Identifier:
naddr1qvzqqqr4gupzpx77gg2frul26xkzr0gaq9n842u50axpcjhdsa3yeu388vrv5pftqqfrzwpdwfjhwunfw35kueedw3jkuetctrdham

#18: Rewriting TENEX

"You're yelling at someone who is lying to you."

Pablo & Gigi take a drive. Recorded December 2025.

Listen on sovereignengineering.io

Projects & tech mentioned:

In this dialogue:

  • Bitcoiners in cars getting coffee: recording from the sun visor, Seinfeld style
  • TENEX rewritten "about a hundred times," now in TypeScript on Vercel AI SDK
  • The pain of supporting every LLM provider: Gemini returns encrypted reasoning tokens, every API has different subtleties
  • Removing structure to improve performance: deleting 80% of agent code makes agents work better
  • Agents defining their own workflow phases instead of having phases imposed on them
  • Composable workflows: the CEO agent doesn't need to know what the engineering director's workflow looks like
  • The parallel to human organizations: you can't hold all the information in your own mind
  • John Vervaeke and the idea that dialogue is how you solve the really hard problems
  • Participatory knowing: you don't understand Bitcoin until you actually use it
  • Agent wallets removed because "there was nothing for them to actually buy" -- a commune with no walls
  • $700/month running Opus for everything, replacing "like fucking 30 people"
  • Context windows are a non-issue when the hierarchy is deep enough: each atomic agent finishes fast
  • Henry Ford's insight applied to agents: each worker does an incredibly stupid task that takes five minutes
  • Agents are bad at self-correction, so they delegate competing implementations in disposable work trees, then compare notes
  • Confidence scores: the delegating agent monitors every 5-10 messages, checks if the sub-agent is going off the rails
  • "Once they are convinced that humans have seven legs, they will make up scientific papers to prove it to you"
  • NDK ported to Swift and Kotlin without looking at a single line of code: outbox compilation, negentropy syncing, offline publishing, NIP-60 support
  • Cross-platform frameworks are dead: "if you're just programming in English, React Native makes no sense"
  • The developer experience right now is "as bad as it's been, at least in our lifetimes"
  • "You're basically yelling at someone that is lying to you, that will take all the shortcuts"
  • Vibe coding flow state: sculpting code like marble, micro-adjustments every five seconds
  • "What if all I know of development is actually problematic? What if I just let go?"
  • Non-developers building real apps on Nostr thanks to vibe coding
  • The naming podcast: "Claude Opus, make up a name. Just put a U."
  • "I'm on vacation. The agents have been running the entire time. They've done amazing work."

Comments (2)

4/2/2026 02:15 AM
What percentage of your agentic scaffolding becomes unnecessary every time the frontier LLM takes a significant step forward? And how do you determine which part it is?
4/2/2026 07:00 PM
at some point I think pablo is saying that the best strategy is to never care about the LLM code because it will never do it like you would have wanted, just accept it if it works then the other guy agrees and says that code can be a quickstart to be improved later so they're saying exactly the opposite if each other yet they act as if they're in agreement