Chief Love Officer at Pragma Research. Connecting curious minds and bridging people together.
Joined February 2026
- Tweets4.9K
- Following201
- Followers416
- Likes1.5K
RT @muellerberndt: I know that OPH sound absolutely crazy. How can spacetime emerge from observation, that's absurd no? But it's the only w…
Crys retweeted
Great week for Cadence. Three releases, and 0.88 lets a brain read several memories at once, learning what to look at on its own. No backprop, no frozen weights, and the builder keeps shipping and sharing what works and what doesn't. Proud to follow this one. 🧠 github.com/muellerberndt/cad… @muellerberndt
$Cadence
#Cadence #OpenSource #NeuroAI
Crys retweeted
If you want to learn more about repair-network theory and its applications in Physics and AI, check out this Notebook.
It can explain everything to you, including why you have subjective experience, how exactly the simulation works and why anything exists in the first place. Get a head-start now, or wait a few years until I get a paper accepted by a mainstream journal.
notebook.google.com/notebook…
Crys retweeted
. @muellerberndt and Pragma Research are working on something that could change the way we think about AI.
Cadence is designed to learn from real experience, not just rely on what it was trained on.
It can keep learning while it works, remember what it learns, and adapt when things change.
That’s a pretty big deal.
Imagine robots that actually get better at their jobs over time instead of needing to be retrained for every new situation.
It’s still early, but the idea behind Cadence is genuinely exciting.
and once Cadence tech actually gets integrated into robots, $70k market cap for $Cadence will look really cheap
Crys retweeted
Cadence v0.83 adds fast-learning, hippocampal-like episodic memory.
github.com/muellerberndt/cad…
Crys retweeted
Another inspirational $Cadence demo for you guys, with source code. Fork it, make a game out of it.
Robots composed of various parts and in-browser Cadence brains that learn from every fight.
arena-brain-showcase.lovable…
Crys retweeted
Did a little evening vibecoding yesterday using Cadence from @muellerberndt.
Please be gentle with my Cadlets:
cadlets.cryptoleszto.workers…
Thank you for the encouragement @crystopher
Crys retweeted
Announcement of an announcement. That's how it's done no?
Btw, don't expect me to keep this up. Stop sending me messages to do xyz so your meme coin pumps. I'll block you. Do something useful: Build a fun Cadence app or even contribute to the project (omg shock!).
Crys retweeted
Hey Cadence community, as Crys just announced, there's a Cadence coding challenge where the winner gets 5 SOL.
Docs are complete so your coding agent should have no trouble putting your idea together. If needed, you can bootstrap small brains on a local CPU/GPU. Remember, there are no frozen weights in Cadence, those brains keep learning.
Vibe-code away 😁
Hey $Cadence community,
I am hosting a 5 Sol Cadence Brain challenge sponsored by @muellerberndt
5 SOL to whoever builds the coolest cadence brain project/demo.
Here are examples: github.com/muellerberndt/cad…
Show us what you got.
Crys retweeted
New Cadence demo shipped today: A brain with eyes that tracks multiple objects in real-time at 40 fps. Biological visual processing is pretty well-researched, we can just copy/paste from nature 😁
Crys retweeted
As a little "thank you" to the @solana community, I built a quick MVP for Mochi, our SOL-colored friend with a bootstrapped browser Cadence brain that learns. I don't have capacity to work more on this particular project but I'm sure that you guys can fork it & turn it into something really awesome, add Cadence token utility and so on. Code is on Github under MIT license.
Also, feel free to implement other ideas! I'm not that creative, I'm sure y'all can do much better than that.
mochi-brain-boost.lovable.ap…
Crys retweeted
For your safety guys, the real Cadence meme coin on @solana is:
dexscreener.com/solana/4tjiq…
Please be careful and double-check if you trade the token. Also, don't fall for fake muellerberndt accounts, I won't make any other announcements in this thread.
Crys retweeted
I'd like to invite all of you to build wild Cadence real-time equilibrium brains! The library is GPL-licensed and well documented. Still way to go on the roadmap but I'd love to see what smart people come up with 🩷
github.com/muellerberndt/cad…
Crys retweeted
FYI, Cadence is now listed on DEX Screener.
Also just release Cadence v0.70.0 which combines animal-type learning (system 1) with recursive cortical feedback (system 2). You can compose brains with a simple "Brain" interface now.
Also, thanks again to the @solana community & @toly for the $3,500 in funding generated so far!
github.com/muellerberndt/cad…
Crys retweeted
I put a Cadence patchnet into a virtual fruit fly. This creates an actual fly matrix in your browser that the fly experiences as a solid world (observation is substrate-independent). All open-source, Github link on the page.
floatingpragma.io/cadence-ex…
Crys retweeted
How the heck does a network of discrete self-reading observer patches output probabilistic quantum mechanics and a smooth 3+1 manifold?
I visualized some of our simulation runs to make it more comprehensible.
simulation.floatingpragma.io…
Crys retweeted
Thanks! We highly appreciate the contributions.
OPH requires specialist knowledge across many disciplines. The simulator hardware and consensus protocol layers are entirely foreign to Physicists while most CS people struggle with the Physics aspects.
I'm not even starting to talk about applied Physics, engineering, neuroscience, biology, philosophy etc. as we're still working on the fundamentals.
Honored to join @muellerberndt on three OPH papers. It was genuinely fun.
The path was hard and complex. Using Oraclizer’s preserve/reflect discipline, I pulled a general layer from one auction fixture and kept pushing until Lean separated four different ways protected behavior can fail.
Bernhard’s willingness to let outside work meet OPH on its own terms stayed with me. I’m grateful for that trust.
The next question feels clear: can the repair process supply the assumptions the proofs need, and what remains impossible under any scheduler? I think that is the path.
github.com/FloatingPragma/ob…
