You love your wife, girlfriend, etc.
You want to avoid your ex-wife, ex-girlfriend, etc.
When you see a platform named X the writing is on the wall.
Machu Pikacchu
npub1r6gg...gmmd
Interested in bitcoin and physics and their intersection.
https://github.com/machuPikacchuBTC/bitcoin
@vnprc I’m reading through to understand ehash a little better and I’m still confused about the tokens.
It mentions they accrue value until some point and then are redeemable for bitcoin. Do they have an expiration? For example, if I have 100 sats worth of ehash at maturity can I hold it indefinitely? Then maybe a year from now (assuming the pool is still active) I can redeem or does it expire at some block height?
Sorry if it’s a dumb question and let me know if there are other docs to read up on.
Redirect
Is anyone working on a DVMCP for code editing? For example, is there a Nostr native way to integrate coding agents into an editor so we can use sats instead of a subscription to Anthropic or Github?
Most people don’t have the hardware to run local models at the speed required for active coding.
#asknostr
Trying to figure out how to connect my Cashu.me wallet with Damus on iOS so I can zap from there. Or Nostur. Nostr wallet connect is setup but zapping still does nothing.
Got my balance to show up in Nostur but can’t send. I don’t want to plug my nsec into Cashu.me which may be adding friction?
Anyone have a magic incantation to make this work?
Something seemingly overlooked in all the Deepseek talk is that Google released a successor to the transformer architecture recently [1].
For anyone who doesn’t know, virtually all of the frontier AI models are based on a transformer architecture that uses something called an attention mechanism. This attention helps the model accurately pick out relevant tokens in the input sequence when predicting the output sequence.
The attention mechanism updates an internal “hidden” memory (a set of 3 learned vectors called query, key, and values respectively) when trained but once training is complete the model remains static. This means that unless you bolt on some type of external memory in your workflow (e.g. store the inputs and outputs in a vector database and have your LLM query it in a RAG setup) your model is limited by what it has already been trained on.
What this new architecture proposes is to add a long term memory module that can be updated and queried at inference time. You add another neural network into the model that’s specifically trained to update and query the long term memory store and train that as part of regular training.
Where this seems to be heading is that the leading AI labs can release open weight models that are good at learning but to really benefit from them you’ll need a lot of inference time data and compute which very few people have. It’s another centralizing force in AI.
1. 
arXiv.org
Titans: Learning to Memorize at Test Time
Over more than a decade there has been an extensive research effort on how to effectively utilize recurrent models and attention. While recurrent m...
No need for new OpCodes to enable covenants.
Write a script that encrypts a method for generating a private/public key pair using the entropy from the last N blocks and can’t be tampered with.
As long as the runtime script is blinded from the machine running it and you can prove that nobody can witness the private key then you can also have the script perform arbitrary computation and verifiably be rewarded with a UTXO.
Ask your friendly neighborhood cryptographer for a prescription for homomorphic encryption today.
#bitcoin #covenants
Are Cashu mints strictly better than FediMints? With Fedi you have the same exact custodial risk because there’s still a single operator for the lightning gateway, but you also have the group of guardians adding friction.
Seems like having all the guardians running their own Cashu mints and spreading your ecash between them all is preferable.
#ecash #cashu #fedi
Block hashes are a good form of entropy that we can all use to sync random number generators.
For example when doing experiments or training AI models. Maybe we can finally fix the reproducibility issue.
In the first half of the 20th century the world had people like Hilbert, Bohr, Von Neumann, Einstein, Feynman, Fermi, Gödel, Noether, Oppenheimer, Erdos, Dirac, Poincaré, etc.
Where are these types of people today? We have Perelman and Wiles. Maybe Maldacena or Arkani-Hamed?
Was the early 20th century a unique period of fertile scientific ground just waiting to be uncovered by smart people or was there an unusually high supply of gifted scientists in that era?