felt (@feltroidPrime) header
felt (@feltroidPrime) profile picture

doing more with less

Magmoor Caverns
Joined August 2010
Photo by felt (@feltroidPrime)Photo by felt (@feltroidPrime)Photo by felt (@feltroidPrime)Photo by felt (@feltroidPrime)Photo by felt (@feltroidPrime)Photo by felt (@feltroidPrime)Photo by felt (@feltroidPrime)Photo by felt (@feltroidPrime)Photo by felt (@feltroidPrime)Photo by felt (@feltroidPrime)
I don't recommend anyone scramble to move their funds to new wallets today. But we should take the risks to cryptography from AI-accelerated math seriously, and minimize our exposure to not just quantum-vulnerable cryptography, but also potentially AI-vulnerable cryptography. The core new area of risk from this viewpoint is, unfortunately, ML-DSA / FHE / lattices. (and it's also another reason, along with quantum, why ECDSA might fall even faster than expected, hence the "fresh address" recommendation) So far most people have been in the mode of thinking "elliptic curves broken, hashes safe, lattices safe". But there is a good chance that the concrete security of lattices will take serious hits from the next two years of AI math. The basic threat model is: factoring is something that naively takes 2^(n/2) time, but over decades smart people have found and optimized number field sieves, and degraded that to 2^O(n^(1/3)), which is why RSA keys and signatures need to be ~400 bytes (and not 64 bytes). What if there are skeletons in the closet like that, both for elliptic curves and lattices, that we are simply not smart enough to discover - but bots soon will be? This is a major part of the reason why for the past year ethereum's lean roadmap has been going in the "hash-only" direction: no lattices, no ML-DSA, no Falcon, no lattice-based commitments inside ZK proofs, etc. Signatures in lean ethereum are all hash-based, either WOTS or SPHINCS-. For signatures and proofs, we already know how to go hash-only. The bigger challenge is for *public-key encryption* - and this goes far beyond blockchains. Secure communication, anonymizing protocols, lots of things need public-key encryption. And unfortunately there are long-standing mathematical theorems showing why public-key encryption cannot be done with hashes alone. You have to have some kind of trapdoor object that has at least one form of usable "structure" - either group theory (incl. isogenies) or lattices or code-based or potentially in the future even more newfangled and spooky things (local mixing?). But for anything that has structure, you should assume that AI will make at least some progress in breaking that structure. Here, one reasonable inference is that if you want to make something plausibly long-term secure, multiply the key sizes by 10. To me that's a very plausible world and something not at all extreme to predict. If AI will bring us 50 years of math in 2 years, then that 50 years of math may very plausibly include a "naive factoring -> GNFS" level of improvement to our ability to break lattices. In that world, lattices will still exist, but they will have to be significantly bigger to guarantee the same level of safety. And at those new larger sizes, hash-based constructions will beat lattice-based constructions on concrete efficiency in every use case where hash-based constructions are possible at all. Theoretically, of course it's possible that hashes are broken too (eg. P = NP would imply that). But I think P = NP is very unlikely. And intuitively, it's much more likely that a mathematical object has exactly no exploitable structure (like hashes are intended to), than that a mathematical object has exactly ~3 forms of exploitable structure (for elliptic curves: associativity, Schoof, pairings) and not some secret fourth form of structure we have not yet discovered that greatly degrades its security (for elliptic curves, ECDLP and pairing security). Similar for LWE, SVP, RLWE and the zoo of lattice problems. For this reason, we do not yet see any reason to worry and start padding the byte size of hashes (if we start to worry more, we would pad the round count first before doing anything to the byte size). Concrete TLDR, my own personal views: * Hash-based > lattice-based, in those situations where hash-based is possible at all * For anything lattice-based, be much more paranoid on param sizes. Remember that blockchains are only a small portion of the cryptography story; this point goes far beyond blockchains and applies to eg. access to websites, secure messaging, Tor / VPNs ... * For privacy protocols, strongly favor NOT putting encrypted notes onchain. Instead, send them offchain through some third-party mechanism. * If it's not difficult for you, keeping your funds in addresses which have not yet been used to make a transaction is a good idea. If it's easy for you, do it. **But be careful about migrations; I personally have lost more money in botched migrations than I have lost in all hacks combined**. * For multisig wallets, doing confirmations offchain is better than onchain, because this way the signatures of signer wallets do not get exposed to the public, so if ECDSA falls to AI much faster than expected, at least the multisig "gracefully degrades" to a 1-of-1 where the 1 is whoever was gathering the signatures - a much better place to be than "anyone can take the money" firefly.social/post/x/210783…
1,246
1,911
892
11,141
3,965,257
Today I call upon the blockchain industry to calmly begin planning for "bunker mode". My personal recommendation is to set in motion a controlled mass migration of assets to fresh addresses, i.e. addresses whose pubkeys remain hidden behind a hash. Holders, starting with large and sophisticated ones, should consider moving the bulk of their funds to addresses that have never signed a transaction. And when they do sign one, they should also move remaining funds to a new address (possibly generated from the same seed phrase). Don't rush. While I believe there is cause for action a rushed migration would do more harm than good. Don't panic either. Moving assets to protected addresses is a simple, preventative step which does not require new cryptography or new wallets. IMO it is now reasonable to brace for the possibility that ECDSA breaks before qday, in the worst case in months not years. By "break" I mean fast private key recovery (e.g. in one week) on available hardware (e.g. a large GPU cluster). Recent days have been humbling for human mathematical intuition. Long-held, unquestioned hypotheses have fallen. This includes the n log(n) bound for integer multiplication and the 3SUM conjecture. In hindsight, May's unexpected disproof of the Erdős unit distance conjecture was our warning shot. Yesterday's OpenAI drop made it clear that mathematical superintelligence is upon us. They say there are weeks where decades happen. We are about to live through weeks where centuries of mathematical progress happen. Could our magic 64-byte ECDSA signatures be too good to be true? Was it just security through obscurity all this time? Elliptic curves feel especially vulnerable to superintelligence. Curves carry rich structure, with room for fancy tricks like Schoof, Frobenius, pairings. (By contrast, hashes are designed to minimise algebraic structure.) Separately, as Ewin Tang can attest, an efficient quantum algorithm sometimes foreshadows an efficient classical one. We should be open to the possibility of a classical counterpart to Shor that breaks elliptic curves and RSA at once. Also noteworthy is the striking under-representation of cryptographic breakthroughs among the 722 mathematical results OpenAI published. I've witnessed first-hand the US government censoring academic quantum cryptanalysis results. Backroom interventionism is my base case. I urge large, sophisticated actors to lead by example. Project11's "risq list" (bitcoin-risq-list.projecteleven[.]com) is a great tracker of exposed BTC pubkeys. Binance, Bitbank, Robinhood, Bitfinex, and Tether have an opportunity to harden their cold storage. Next month I'll address institutions in London in a live Q&A (forum.ethereuminstitutional[.]org/london-2026). Again, please do not rush. Wallets holding under 50 BTC enjoy partial cover from "Satoshi's shield", i.e. his 20K exposed addresses that hold 50 BTC each. Load-bearing signers like oracles and L2 security councils should consider rotating ECDSA pubkeys with every signed message and/or multi-signing with a hash-based schemes like SPHINCS. Exiting bunker mode safely will require post-AI cryptography. My inclination is to go all-in on hash-based cryptography and avoid structured mathematical assumptions entirely, whether from curves, lattices, or isogenies. A single battle-tested hash (e.g. from the SHA or BLAKE families) yields plausible post-AI security. The Ethereum roadmap on strawmap[.]org fully embraces hash-based cryptography with end-to-end formal verification as a response to the quantum threat. Those timelines must now be revisited and accelerated in light of mathematical superintelligence. I'll be pushing for maximum defensive acceleration.
5,955
5,503
2,295
23,111
6,709,657
for repro: I gave it a py kernel just like they did (which we had for about 6mo now btw)
2
1
1
146
25,778
Flock is the fastest SNARK prover ever built. But how much faster can it get? snark.fast is an open contest to find out. Already 100+ submissions in and the current record stands at >1.2 million BLAKE3 compressions/s. Co-organized w/ @SuccinctLabs @EigenLabs & @ethereumfndn This is what good cryptography research looks like: publish your best work, then invite the world to improve on it.
5
7
3
35
9,531
felt retweeted
for the last few months i have been working on long running persistent agents. i believe this is the next paradigm shifting product after chatgpt and codex. there are a number of challenges, and i'm currently writing a longer post with some learnings & thoughts on why i believe the next big unlock is solving theory of mind loose thoughts: - the next paradigm-shifting product after chatgpt and codex is the persistent personal/work agent - the blocker to making such agents is that our models can't "read the room" - reading the room sounds soft but it's deep theory of mind - reading the room is the difference between talking to a bot and talking to an embodied identity you can trust to accomplish tasks and communicate with your colleagues - reading the room an understanding of who "your person" is, their desires and motivations, and who your audience is - concretely, it's tracking what i know vs what my user knows vs what the room knows. information asymmetry as a first-class skill - part of this is understanding compartmentalization. enterprises call this tenting and workspace isolation. "if my user is part of this tent, don't divulge information in a broad channel that's not also tented" - if i give a friend my home address, i trust them not to announce it in a room of a thousand people. I do not yet trust a model to exercise that same discretion. - harnesses like openclaw keep trying to solve this at the application layer with permissions, workspace isolation, context injection, and increasingly elaborate agent harnesses. necessary for now, but this likely futile as the final answer. discretion has to live in the model - side effect: this is also part of why model writing reads as slop. good writing is meeting your audience where they are. same missing skill tldr: persistent agents live or die on reading the room, and theory of mind may be one of the last major gaps before agi
63
34
19
593
64,261
Dario and @AnthropicAI tonight... Nobody will notice a little rename, right?
119
186
50
5,096
741,110
can someone in anthrophic just start torrenting the weights
50
128
14
4,870
203,646
Replying to @Koviett @pubity
The difference is with the time we’ll invest in making the tools we need and they’ll be open source and as good as Microsoft’s and it’ll benefit everyone. We stop being someone else’s bitch and give to the world at the same time.
1
1
257
🚨 Shocking: Frontier LLMs score 85-95% on standard coding benchmarks. We gave them equivalent problems in languages they couldn't have memorized. They collapsed to 0-11%. Presenting EsoLang-Bench. Accepted to the Logical Reasoning and ICBINB workshops at ICLR 2026 🧵
149
272
224
2,136
1,266,548
1/ The STRK20s Technical Paper is now live. It outlines the design, architecture, and core ideas behind STRK20s, a privacy capability for any ERC-20. If you’re bullish on privacy as crypto’s next major unlock, this paper is for you 🧵
59
116
57
386
167,651
Added support for RSA-2048 signature verification. It uses the 384 bit modulo builtin to its fullest by using a residue number system and the Chinese remainder theorem across 11 comprime channels. garaga.gitbook.io/garaga/usi… Gas costs are 2x more expensive than EC-based sigs.
2
2
10
611
Replying to @GrapheneOS
We can sponsor you. Please contact us and share the specs to [email protected]
104
119
44
2,816
63,044
privacy pools is now live on @Starknet deposit ETH, USDC, or STRK, withdraw to a fresh address, and use any app on starknet privately without revealing your onchain history or public identity starknet is the place for privacy
30
41
34
193
57,692
felt retweeted
I am soooo excited to have @feltroidPrime join me in this important mission that the fight for Freedom Tech. This dude is crazy, but in a good way. The kind of crazy that can self learn ANYTHING, like really. He was able to grasp super moon math crazy concepts, and implement super hardcore low level optimizations for deep elliptic curve techniques and other crazy stuff, and to ship them for production usage. Now he is entering the agentic era (it's interesting cause he was in the denier / sceptic camp for a while on this!). It's a signal to me! And man, if this dude puts the same GRINTA for grasping agentic wizard stuff ( and he will!), no one is ready. Anyway, warm welcome to StarkWare!!! FGU!!!
after «Fixing My Life In a Day» for the past 6 months I’ve officially joined @StarkWareLtd with Abdel’s team to work on freedom-enabling and privacy-preserving techs in the current and upcoming agentic era. Grateful for this opportunity ; waking up excited which is a blessing!
16
10
3
71
5,289
felt retweeted
Replying to @rauchg @vercel
Agreed but it’s time to go back to fundamentals in terms of security aspects too. We need to raise the bar right now it’s euphoria time and we sacrifice a lot security practices for convenience. It’s time to prepare solid foundations for the agentic era. Cryptographically signed skills / tools / prompts, strict versioning, modularity and strict isolation of sensitive components etc
1
2
1,256
Replying to @dimethyeth
@Shade_L2 is a scam, they haven’t built anything and they won’t deliver shit. They just surfing on the current privacy narrative to grab a list of emails and accounts of potential buyers, I suspect they will soon launch some kind of ICO since it’s also the current narrative. @yunaintern is promoting it a bit too much for obscure reasons The article only talks about vision but the promises are simply impossible for an anon team with a X account created in Dec 2025 to ever deliver. Everything from the article to the website and docs is fully vibe coded with 0 root in technical reality. Yet big reputable accounts are promoting it or acknowledging this newcomer, or even congratulating it. Don’t loose your critical sense guys. This is an obvious scam.
You can clearly sense this article was written from a LLM prompt optimizing for marketing and aspirational promises, rather than a prompt from marketing grounded in any existing document, paper, or implemented technology.
1
1
7
687
@NoirLang, Groth16, @SuccinctLabs, @RiscZero, @drand_loe and more primitives are all available securely and at scale. Documentation: garaga.gitbook.io/garaga/ What will you build or bring to @Starknet ? 👀
1
7
1
21
2,302