PiSwap L2, aka PiEthereum Wallet v2, a little walk through of the new functions. I will post a demo of the any token swapping function which now utilizes @Uniswap v4 Universal Router. This is a free open source raspberry pi python project for any L2 EVM chain. Build it yourself via instructions on github. Link in Bio. Touch the blockchain with no intermediaries and many private keys.
2/ Breadventure is now available to play on Steam and it’s truly a delight! It’s also on sale for a limited time, so make sure to purchase it soon!
https://t.co/z2nTTpVFjq
says something that AI centered games aren't performing well. few are interested in observing a simulation. most people want to be the actual main character.
and AI centered games of chance won't hit the dopamine receptors of the end user
this one goes out to the base:0x9f86db9fc6f7c9408e8fda3ff8ce4e78ac7a6b07 homies
if you have staked and earned conviction
I'll give you a wedgie!
https://t.co/qXfufztPgi
I have a major result to share.
I’ve found conclusive forensic evidence that the top 4 Chinese AI labs—DeepSeek, Moonshot, MiniMax and Ant Group—have built their newest models on NVIDIA’s most advanced Blackwell chips, which are barred from export to China.
This sounds insane, but I discovered that almost every AI model carries a hidden signature of the GPU that trained it. When a model is first set up, the starting values of its weights are laid out in a pattern that depends on the exact GPU. Each combination of GPU and seed number produces a unique pattern—a sort of fingerprint of the hardware used to create the model. Traces of that pattern survive training into the published model.
Again, this evidence isn’t based on any insider information or leaking. It’s based on a forensic analysis of the model weights released by the companies themselves.
I searched more than six trillion combinations of chip layout and seed number and matched them to open-weights models. In doing so, I found conclusive evidence that six Chinese AI models—including DeepSeek-V4-Pro, Kimi K3, and MiniMax-M3—were initialized on NVIDIA Blackwell systems. Four of the models match Grace Blackwell, the GB200 and GB300 parts in NVIDIA’s flagship liquid-cooled rack-scale systems.
I then verified these results on real hardware. Running unmodified PyTorch with the seeds I’d recovered, a Blackwell B300 reproduced DeepSeek-V4-Flash’s untouched starting weights bit for bit, and a rented GB200 reproduced Kimi K3’s to a correlation of 0.999995. A Hopper H100 reproduced none of DeepSeek-V4-Flash’s, and changing the seed by one broke the match. (DeepSeek’s seed is 953049583 by the way.) The method also was verified to match 24 known open-weights models to the hardware that developers said they used.
It is exceedingly unlikely that these results are a coincidence. Across six trillion combinations, the best accidental match should reach only about 8 sigma (i.e., eight standard deviations above what random chance would produce). The real matches range from 49 to 1,385 sigma, and even the weakest single test hits 20.
These were data-center-scale training runs. With some analysis of release schedules and model details, I estimate that the largest, Kimi K3 and DeepSeek-V4-Pro, used 2,000 to 3,000 Blackwell GPUs each. Clusters of this size cost on the order of $150 to $200 million each to buy, and across these labs alone would add up to more than half a billion dollars of export-controlled GPUs, not counting experimental compute or otherwise. It’s quite likely that we see multi-billion dollar expenditure on export-controlled GPUs.
I am highly confident in this result, and it has been independently verified by senior researchers at OpenAI and Anthropic. You can read the full report at https://t.co/YOtzND0Jqk.
There is much more I’ve identified about these models through other forms of checkpoint forensics like this, and will have more to share soon.
@latimes if only they got a piece of the 4 billion dollar proposition check the other california non-profit immunology research institute just got from california voters
IMPEACHMENT, Article III
The law requires our military to withdraw from hostilities within 60 days if Congress does not authorize the activity.
Operation Epic Fury has been underway for 256 days. Congress has not authorized it.
Hegseth is breaking the law.
Not even debatable.
so if we want ethereum to win (first we need to survive, tbh lol), we need to fund its security. there are no shortcuts here guys. onchain security is everyone's problem, not someone else's job. TheDAO Security Fund (where i'm a curator, so yes, i'm totally biased) has a number of open security initiatives that deserve your attention and, more importantly, your funding. if your protocol depends on ethereum security work getting done, put in 10-15% and let TheDAO Security Fund help raise the rest. it's a lot cheaper than funding the entire thing yourself. and pledges are only collected once an initiative is fully funded. no free-riding. no money collected for initiatives that don't make it.
my personal top two picks are (remember i'm biased obviously so go thru them by yourself :)):
- formally verified Vyper compiler (https://t.co/KWTWHJYOqX)
- production-ready local-first Safe UI (https://t.co/rKSLwauzpj)
ethereum's security won't fucking fund itself. stop treating it as somebody else's problem. put some skin in the game.
https://t.co/J8iQLlK4IL
PS: info on this second round https://t.co/Sh7yWdqunH
tldr on how fair launches work:
• first 60 sec, everyone buys at one price
• nobody gets more than 25% of the curve
• starts ~$18k mc, graduates ~$55k mc
• 20% sniper tax after graduation, gone in 5 min, burned
• creator earns 1% while they hold
• creator sells 10%+, their fee gets burned
• no token tax, no owner
Built so the people who show up first don’t just get farmed.