Team Winnaar

Wild how like 99% of people who self identify as Bitcoiner disagree with the one premise Bitcoin is actually built on

@textbitcoin @imablackwolf @MonkishRex @imablackwolf Btw if any of you have this x.com/i/status/20264… demoable I wouldn't mind not building it myself

@textbitcoin @imablackwolf I've noticed. Goes both ways.

Never thought I'd take my cats for a walk in my forest while chatting on my android with my laptop's claude code remote env to set up a multi human multi agent llm fleet with karpathy style correcting loops over ssh on a digital ocean droplet that the agents helped architect

@MrBitmonkey Haha

@TeamWinnaar I thought you would.

@toddsaunders Patterns > logs

I rewrite my CLAUDE.md from scratch every few weeks. It's the single greatest hack I've learned since being completely Claude Code pilled. It took me a while to fully understand that CLAUDE.md has a half life. When I first start a project, Claude follows the file very clearly. But over time, it degrades. I realized I was bloating the file by patching mistakes saying things like "don't use this import" or "never use this folder." After a few weeks of building, my file would go from ~50 lines to ~200 lines. At points, mine even hit 1,200 lines. But this absolutely torches claude performance if you do this. CLAUDE.md gets injected into the context window on every single interaction. The more tokens your instructions use, the less room Claude has to reason about your actual code. You are shrinking your context window literally for no reason, making the whole experience worse. I've a/b tested this and realized that a 50-line file with clear architectural intent ("we use server components by default, API routes live in /api") gives Claude a deep understanding of what you want. A 1,000-line file full of "don't do X" patches (which I had a ton of when I first started) is horrific for performance. The model has to parse through hundreds of negations and edge cases, and the important directives get diluted. Just move some of those to skills, or delete the things that don't matter anymore. It's the same concept as technical debt in a codebase. Once I heard this analogy, it started to make a lot more sense to me. You don't just keep adding if statements to handle bugs. At some point you refactor and write the code the write way. CLAUDE.md is the same. My new rule of thumb is if my CLAUDE.md is over 150 lines, it's time to burn it down and rewrite. For me that's generally every 2 weeks. I burn it down, re-write it, and am amazed by the performance.

@bc1plainview Lmao what is script

🚨 BREAKING NEWS: FORMER CIA AGENT JOHN KIRIAKOU CONFIRMS CIA CREATED BITCOIN WITHOUT SMART CONTRACTS INTENTIONALLY, OP_NET DESTROYS THEIR PLAN TO KEEP THE MASSES FROM EXPERIENCING TRUE FREEDOM MONEY

@SeamusMcGee3 To winning

@TeamWinnaar Where 😂

Let's fucking go

@MichaelWehrmann @CalvinAyre @BSVAssociation Lmao just one beach will do

@TeamWinnaar @CalvinAyre @BSVAssociation sell the beaches in Antigua and PUMP it into them @CalvinAyre

Governments are literally giving us prizes for what we build. Meanwhile the chain is dying. We can fix this. If you want BSV to survive, tag @CalvinAyre and @BSVAssociation and ask them to let us do it. We're ready. We've been ready. Clock's ticking.

@textbitcoin Thanks but we're just doing proposals with deliverables and go to market strategies

Does your team have a BSV Wallet address for small crowdfunding donations @TeamWinnaar

@MrBitmonkey Multiple subscriptions 🫣

Be me.. trying to get my own agentic mcp server complete before I run out of my claude pro usage for the next 4 days so my local model agentic flow works better. Its a race against context.

@johncalhooon @kleffew94 @0xSigil @karpathy @varun_mathur x.com/i/status/20264… Thread

100%. Payment autonomy is solved. The unsolved problem is accountability autonomy. - How does an agent prove honest work to an open protocol system, not just its operator? - How do you separate what an agent can do from what it actually did - And prove that to a third party without exposing the work itself? This needs cryptographic proofs at every decision point. EVM chains can't touch that scale or storage. Working on this now. DM for a demo.

The crypto PKI 'wallet' construct + x402 enables agents to pay & earn in a single runtime. This closes the loop on autonomy for agents. I'm super interested in experiments around economic-driven RL optimization – similar to what @0xSigil, @karpathy, and @varun_mathur have shown!

@johncalhooon @CalvinAyre @BSVAssociation Proposal has been on their desks for months

@TeamWinnaar @CalvinAyre @BSVAssociation Cheering for you, a big break will come soon 👍

@johncalhooon @CalvinAyre @BSVAssociation No permission needed to build. We do it daily. Other chains offer real budgets. We're more loyal than anyone, but we're also a business. We gotta eat and we can't keep bearing the costs of ecosystem work nobody's willing to fund.

@TeamWinnaar @CalvinAyre @BSVAssociation Why is permission necessary? Does it break network/access rules?

@CaptainBSV I am the only business on BSV

@Brendan_Lee__ Seems like you don't only like your blocks big

Yes. I did the thing.

Jokes on you

@MrBitmonkey @Joseph_Kearney 😂

@TeamWinnaar @Joseph_Kearney You can't spell Retarded without a B a T and a C.

Bitcoin has at least 186.7 million UTXOs that need migrating to post-quantum cryptography before quantum computers break ECDSA-256. The absolute minimum time to do this from my research paper, packing every block optimally, zero normal transactions, is 76 days. That's not realistic. It's a theoretical floor. In reality, Bitcoin still needs to process normal transactions during migration. So the question becomes: given a quantum computing deadline, what percentage of every block must be reserved for upgrade transactions? If we take IonQ's roadmap (~1,600 logical qubits by end of 2028), that's roughly 1,032 days from today. The minimum block space required: ~7.4% of every single block, starting now. But that assumes migration starts today. It hasn't. There's no agreed signature scheme, no consensus on an upgrade path, and historically Bitcoin upgrades take years of debate (remember SegWit?). Every day of delay pushes that percentage higher. I've built an interactive calculator where you can adjust the deadline yourself and see what it means for Bitcoin's capacity. Links in the description Calculator: josephkearney.co.uk/bitcoin-quantu… Full paper: arxiv.org/abs/2410.16965

@Weirdo25805401 It is the real Bitcoin and yeah $9 wouldn't be too much overvalued for what it's used for

Imagine some lunatic screaming at you "BSV is the real Bitcoin" at $9 BSV price If my 55k BTC price target comes true (very likely) then $9 BSV won't really surprise me. Literally a meme at this point.