Pinned Tweet
Rust needed its FastAPI moment. So I built it.
fastrs — validation, OpenAPI docs, and routing that generate themselves from your types. Zero runtime cost, because it all happens at compile time.
I documented fastrs, my FastAPI-style framework for Rust, with @thallyio .
Thally read the repo and opened a PR with 15 pages in less than 10 mins 🤯 . I reviewed it, merged, and it was live.
Docs: fast-rs.thally.app
How I did it (and what broke):
mahmudsudo.hashnode.dev/publ…
rust fanatic retweeted
Interesting note: CC agents are really good at narrow tasks, but completely fail if you stray from the path you have them on. Example:
Research task --> slow --> optimize --> research task
vs
Research task --> slow [spawn new agent, optimize] --> research task.
This increases the chance we see "agent teams" increase in popularity! Check this tweet in a few months
rust fanatic retweeted
Something I see nobody profiling yet is the time it takes an agent to implement something in various frameworks/languages. Def some alpha here.
rust fanatic retweeted
After 3 years of hard work, we are happy to release Zippel, a Language, Compiler, Runtime, and Analyzer for ZK Proof Systems (github.com/eniac/zippel). The paper is here (eprint.iacr.org/2026/2517). Zippel will appear at S&P 2027. 1/5
rust fanatic retweeted
Hello World 👋🏼
Today I’m excited to announce @thallyio, the product knowledge layer for software teams.
When your product changes, Thally finds the affected docs, website and support content, then opens evidence-backed PRs with proposed updates for your team to review and approve.
Not really. Competitive prover markets don’t automatically centralize building, hardware + recursion keep getting better for local proving, and lighter verification is what actually helps more solo stakers stay online. “Slow Solana” is a stretch.
I think the later is the best path
And I honestly feel in the likely breaking of the ecdlp , new and harder assumptions will rise to take its place
As a researcher, I like to think about what-if scenarios, so let's consider one here: what if there is no more public-key encryption?
Public-key encryption is used in countless ways in our daily lives, including online shopping, banking, and protecting anything we want to keep private from the rest of the internet.
What if the mathematical problems that underpin public-key encryption (factoring, discrete log, lattice problems, decoding of codes) are all easy to solve because of mathematical breakthroughs by artificial intelligence or quantum computers? What happens to internet commerce?
I see two worlds:
1. We rely on trusted intermediaries to send us keys for encryption. There is no more public-key encryption, and all we have is symmetric encryption like AES. In this case, we need trusted third parties to send counterparties dedicated keys for encrypting.
2. A second possibility that I like better relies on hash-based zk-STARK systems to create digital signatures that depend only on the hardness of hash problems.
In this world, you could still have payment systems and internet commerce for sending payments, but you'll need to solve the challenge of easy, direct encrypted channels between counterparties.
My tl is either Ai doomsday reviews or Ai based cryptography reports
Can we go back to the old days 😭?
I think he might be wrong about so many other things
fun fact: satoshi refused to use the NIST seeds probably because he was somewhat of a bunker mode retard himself, instead used the more obscure secp256k1, and most other cryptocurrencies followed
this is why we can’t directly use industry standards like passkeys in blockchains, and instead we ship people purpose-built hardware wallets made by fucking amateurs, leading to routine 9 figure hacks
anthropic’s discovery reveals that in retrospect his panic wasn’t necessary and we could’ve just used the standards
thanks bunker mode retards!
(this comment is exaggerated for comedic effect but it’s not wrong)
The only thing stopping @drakefjustin concerns from being a reality is finding a classical version of shor’s theorem
Time to become @paystack devrel
With this same line of thinking , we should be ok with dangote having monopoly over cement and petroleum .
Stop this faux intellectualism
Competition breeds efficiency and growth
Monopoly breeds mediocrity
When you talk about it, they hit you with the gimmick of “the sky is enough for everyone to compete.”
Lol, bro, you ain’t competing you just wish the attention was on you, not her.
Then they say there are so many social media platforms.
Yeah, but none are one and the same. They copy features, but X differs from Facebook, just as IG does from LinkedIn. Don’t even mention Snapchat, which I still don’t understand, or TikTok and YouTube.
