eprdude.eth ๐ง๐ฆ๐๐ ๐ก๏ธ retweeted
We can't let them turn 'open models' into 'Chinese models'
US TSY. SEC. BESSENT: ANY AI-BASED ATTACK YOU SEE COMES FROM CHINESE MODELS
Readers added context they thought people might want to know
The most notorious of recent cyber attacks, was carried out by a US model against hugging face.
openai.com/index/hugging-โฆ
eprdude.eth ๐ง๐ฆ๐๐ ๐ก๏ธ retweeted
AI threats are real. @ethereum validators are in danger. It has never been easier to steal validator private keys and slash validators.
To address this, @LidoFinance contributors developed github.com/lidofinance/tee-sโฆ. The only TEE-native signer for Ethereum validators!
eprdude.eth ๐ง๐ฆ๐๐ ๐ก๏ธ retweeted
Open-source is essential. I built MongoDB, one of the worldโs most popular open-source databases. Now Iโm building Viam, an open-source harness for physical AI.
I donโt want a world where millions of developers are dependent on the paywalled technology of just a handful of companies. No one should.
I want to see 1B+ new physical AI deployments over the next ten years.
Hereโs why open source is essential to getting there. ๐งต
Readers added context they thought people might want to know
As co-founder and CTO, Eliot Horowitz was a primary architect and public advocate of MongoDB's transition away from open source to its current SSPL source-available commercial licensing model.
opensource.org/blog/the-sspl-โฆ
eprdude.eth ๐ง๐ฆ๐๐ ๐ก๏ธ retweeted
Earliest public entry point for the most valuable companies on earth being $1-2 trillion is mega cooked
Good news is DeFi has been building the solution for the past 8 years
eprdude.eth ๐ง๐ฆ๐๐ ๐ก๏ธ retweeted
We're teaming up to accelerate open-source Al.
OpenRouter is now integrated into Hugging Face directly. Same token, same endpoint, plus live model stats on the Hub and an inference widget we redesigned together.
The best of open, in one place. huggingface.co/blog/openroutโฆ
eprdude.eth ๐ง๐ฆ๐๐ ๐ก๏ธ retweeted
๐๐ป๐๐ฟ๐ผ๐ฑ๐๐ฐ๐ถ๐ป๐ด ๐ก๐ฒ๐๐ต๐ฒ๐ฟ๐บ๐ถ๐ป๐ฑ ๐๐ถ๐ด๐ต๐ ๐๐น๐ถ๐ฒ๐ป๐
Trustless RPC for every wallet.
Verify Ethereum yourself, without running a full node.
๐๐ผ๐ ๐ถ๐ ๐๐ผ๐ฟ๐ธ๐
โ One trust anchor: a recent finalized beacon block root
โ Consensus: sync committee updates over libp2p, with signatures and Merkle branches verified
โ Execution: state on demand through SNAP, every range proven against the state root
โ Performance boosted through partial state caching for selected and frequently queried accounts.
๐๐๐ถ๐น๐ ๐ผ๐ป ๐ฏ๐ฎ๐๐๐น๐ฒ-๐๐ฒ๐๐๐ฒ๐ฑ ๐ก๐ฒ๐๐ต๐ฒ๐ฟ๐บ๐ถ๐ป๐ฑ ๐ฐ๐ผ๐บ๐ฝ๐ผ๐ป๐ฒ๐ป๐๐
โ eth_call and eth_estimateGas run locally in the Nethermind EVM, nothing forwarded
โ Developed inside the Nethermind client codebase
โ Glamsterdam-ready
๐๐ผ๐ฟ ๐๐ผ๐, ๐ถ๐โ๐ ๐ท๐๐๐ ๐น๐ผ๐ฐ๐ฎ๐น๐ต๐ผ๐๐:๐ด๐ฑ๐ฐ๐ฑ
Point MetaMask, or any wallet, at it.
In active development. Congrats @__flcl and the team.
Follow the progress and share feedback:
github.com/NethermindEth/netโฆ
eprdude.eth ๐ง๐ฆ๐๐ ๐ก๏ธ retweeted
We've come a long way in a short time for formal verificationโone of our strongest tools for putting the S in CROPSโto become a new normal. Here, verity formally verified EIP-8282 and in doing so found real discrepancies for researchers to address.
Thanks for the work @Veritylab
We formally verified EIP-8282 for the @ethereumfndn.
Proved directly on the bytecode, checked by Lean.
veritylabs.dev/our-work/eip-โฆ
eprdude.eth ๐ง๐ฆ๐๐ ๐ก๏ธ retweeted
Replying to @VitalikButerin
@VitalikButerin just opened Devcon and set the tone for the week.
A formally verified OS, open silicon, self-sovereign healthtech, free discourse, private communication, trustworthy AI โ these are all part of Ethereum's mission.
"We need to be able to trust our operating systems.
We need to be able to trust our hardware.
We need to be able to trust the things that are most relevant to our health.
We need to be able to trust our sources of information and connection to other people.
We need to be able to trust our communication tools.
We need to be able to trust our AI.
What is the answer? We need to figure out together." - @VitalikButerin
๐ Devcon 8, Opening Ceremony
everyone believes intelligence has unlimited demand, not just the maxis. but what the maxis get wrong is that the intelligence anthropic and openai and frontier models are building isnโt the generic kind youโre picturing. let me explain-
the two biggest uses of compute are crudely either training or inference. the โgeneric intelligenceโ everyone says has unlimited demand comes from mainly inference, and inference needs very different power and utilization than training. inference comes in bursts, so you have to by default overbuild capacity to handle the peaks until the utilization problem is solved, which means almost by definition we are going to overbuild. but training doesnโt have that problem: the workload is steady and the demand curve is knowable. so data centers end up overbuilt right now precisely because utilization isnโt optimized yet.
so anthropic and the other frontier labs are building their future on training, which is why they are saying they can soak up all this excess capacity. but their โintelligenceโ isnโt workplace productivity. itโs the leap toward singularity-type outcomes: new drug discoveries, cyber capabilities, recursive self-improvement. these are true 1-of-1s. thatโs why when anthropic talks about โexistential threats to humanity,โ it isnโt talking about job losses, not really. itโs saying theyโre on the precipice of unbelievable discoveries, because theyโre still TRAINING MORE MODELS.
but distillation is real and open source is real. the half-life of any frontier model is tiny because nothing protects it the way IP protects a patent. imagine a rare disease drug company that spends billions on research, and two months after a breakthrough a generic shows up. that company can never recoup its investment, unless it can discover an infinite number of rare disease drugs. THAT is the kind of intelligence anthropic is betting on. thatโs what the $800bn of compute spend is betting on.
so the questions you need to ask are: 1) if intelligence is two-tiered, what does each tier cost? whatโs the cost of AGI, and whatโs the cost of the generic intelligence that has infinite demand? and 2) are we allocating capital with the right risk curve and decay function to find out whether the AGI payoff is real? and if it is real, could a private company ever truly own it?
thatโs the cognitive dissonance at the heart of this entire paradox: everyone is gaslighting each other about what intelligence is as the unit of measurement is purposely kept vague by both the users and the suppliers. generic intelligence has unlimited demand, but itโs a certain race to the bottom. "superintelligence" is the opposite: it can't have unlimited demand by definition, but if it delivers even half of what it promises, no private company will ever get to keep it, at least not without the government stepping in to guard the moat (which is why anthropic keeps playing this angle up, without having shown any demonstration they are near this level of discovery at all). and at that point you have to ask who is going to hold the bag, and whether that was in fact, very intelligent at all.
I've become inceasingly bearish on AI / AI capex in recent months. I've worked hard to speak to VC's and founders to understand where I may be wrong. I've ultimately concluded the AI maxis believe intelligence has unlimited demand and understand very little about microeconomics or accounting.
It is eerily reminiscent of Covid when everyone clapped like seals when Fauci told lie after lie, including the double masking idiocy. We are well past that point in the AI trade. The beauty of it is, the AI maxis have chosen to lever up to the max and take out hundreds of billions of dollars of debt based largely on vibes. The die is now cast, and there is nothing anyone can do about it. Now we just sit back and watch the AI trade collapse just like Leo did.
eprdude.eth ๐ง๐ฆ๐๐ ๐ก๏ธ retweeted
Excellent overview of Aave from @Citi Research.
Citi positions Aave as the collateral mobility protocol that will make trillions of dollars in assets productive as they move onchain.
If you're new to Aave, this report offers a great overview of where the future of finance is heading and how Aave is building its backbone.
New @Citi research positions Aave as a collateral mobility protocol for tokenized capital markets.
It calls collateral mobility one of tokenizationโs most important use cases, with Aave able to take on functions that move tens of trillions of dollars in traditional finance.
eprdude.eth ๐ง๐ฆ๐๐ ๐ก๏ธ retweeted
Realtime open video generation keeps getting faster. FastH3-Omni is out with weights + training code, and vLLM-Omni serves it day 0 โ including on consumer RTX GPUs. Congrats to the @haoailab team! ๐ท
As promised, FastVideo open-sources @MiniMax_AI FastH3 Omni, training code included! ๐
- Real-time gen @ 12x faster on NVIDIA Blackwell
- NVFP4/FP8 for Spark + RTX!
- ComfyUI day-0
Collab with @NVIDIAAI , Nuva Lab @nuvalab and Reactor @reactorworld
Open source must win! โฅ๏ธ
eprdude.eth ๐ง๐ฆ๐๐ ๐ก๏ธ retweeted
We want Ethereumโs rollups to work as one network. Thatโs why weโre building @etheconomiczone support into the Nethermind Client.
teapawt.net/svlachakis_/status/211โฆ
We successfully tested @Nethermind's standalone Rollup0 prover with the @etheconomiczone composer, including L1 proof acceptance and restart/recovery.
One step closer to a public EEZ testnet with multiple prover clients!
WIP draft: github.com/NethermindEth/netโฆ
eprdude.eth ๐ง๐ฆ๐๐ ๐ก๏ธ retweeted
Announcing FlashMap: a new speculative decoding algorithm based on flow maps! >6x faster inference on Qwen3-8B ๐งต
Project page: jshen.net/flashmap
arXiv: arxiv.org/abs/2610.24365
w/ @Jay41969851 @_ryantlee @jonathanmay @cosmo_shirley
eprdude.eth ๐ง๐ฆ๐๐ ๐ก๏ธ retweeted
Nethermind Client 2.2.0 is out. For RPC providers: faster responses that use less memory, and faster tracing on Glamsterdam blocks.
Mandatory for Hoodi nodes before the Glamsterdam fork on Oct 26. github.com/NethermindEth/netโฆ
eprdude.eth ๐ง๐ฆ๐๐ ๐ก๏ธ retweeted
My electricity bill for my primary residence was 24% higher in Oct. 2026 than it was in Oct. 2025. It's the same house with all the same appliances. This isn't what happens in an economy where the inflation rate is anywhere near 2%. That's just one bill. They're all going way up.
eprdude.eth ๐ง๐ฆ๐๐ ๐ก๏ธ retweeted
The growth curve of Venice.ai users:
1st million: ~13 months (May 2024 โ June 2025)
2nd million: ~7.5 months (June 2025 โ early Feb 2026)
3rd million: ~3 months (Feb โ April/May 2026)
4th million: ~3 months (May โ Aug 13, 2026)
5th million: ~2 months (Aug 13 โ Oct 19, 2026)
base:0xacfe6019ed1a7dc6f7b508c02d1b04ec88cc21bf #AI






























