Joined August 2015
- Tweets81
- Following186
- Followers486
- Likes500
Duc V. Le retweeted
We're now accepting applications for research interns at @a16zfrontier, for summer '27 in NYC! Application deadline is Nov 20th. You'd be working with me @skominers @SuccinctJT @ittaia @PGarimidi @jneu_net @ErtemNusretTas, as well as other interns, visitors, and our advisors (such as @danboneh).
Our interests are broader than ever---ambitious undergrads with a keen grasp of emerging technologies, grad students looking to transition to building and entrepreneurship, as well as PhD students pursuing a more traditional academic career.
Research areas of interest include AI and machine learning, agentic commerce, cryptography, security, distributed computing, hardware design, economics (micro, macro, and empirical), energy, material science, biology, incentives, quantitative finance, market and mechanism design, foundations of computing, and pure mathematics.
For more details and to apply, see jobs.a16z.com/jobs/andreesse…
1/ Agents paying agents needs reputation that fake accounts can't game. Our new paper (collaborated with Reza Ramji, @PrestonVanderV, and Ziming Zeng), Anduin, builds it from payments rather than reviews (arxiv.org/pdf/2611.05586).
3/ We show that an attacker's influence on a buyer's ranking is bounded by the money the attacker pays to providers that buyer already uses. Creating fake accounts costs nothing; moving a ranking costs real payments.
4/ This work aims to fulfill the trusted discovery gap highlighted in Circle's Building the Open Agentic Economy vision.
nitter.gq/circle/status/20879321…
The agentic economy is starting to take shape.
Circle’s new vision paper explores an open economy where agents can provide labor, be discovered and ranked on merit, and get paid for the work they perform.
A new internet primitive is emerging: software that can do work and earn when that work is used.
circle.com/blog/building-the…
Duc V. Le retweeted
The age of agents is going to be amazing for most programmers, but it'll be hard times for the cantankerous ones.
Specialized knowledge and rare skills will no longer offer immunity for those who are a pain to work with.
Best learn some basic people skills.
Duc V. Le retweeted
"Hi Claude good morning how are you I'm great thx build me a YouTube-style site but restricted to videos that have been repeatedly recommended on HN over the years pls. Don't make any mistakes."
Duc V. Le retweeted
Tell Sam Altman that I use my 64GB of RAM to run print("Hello, World!"). I want him to know it's me.
Duc V. Le retweeted
In 2 weeks, I'll present at Ethereum's Devcon! 🎉
I'm discussing how to Handle Private State Bloat, joint work with @levduc @pedrorechez Zhipeng Wang
1/ AllowList (e.g. UTXO) has consensus number 1. DenyList (nullifier list) doesn't -- non-membership needs agreement. Private payments are both at once, weilded by one spend proof.
That pairing is what buys unlinkability, and what makes both sides seem unprunable. @PrestonVanderV on how to prune them anyway.
Can privacy scale without becoming its own bottleneck?
@PrestonVanderV of @circle is joining Devcon 8 to tackle one of privacy infrastructure’s hardest problems: growing private state.
Catch his talk in the Privacy & Consent track on nullifier pruning & the privacy-scalability trade-off.
2/ The usual reading is that you're over-paying consensus on the AllowList side. The better reading: you already bought total order for the DenyList, so use it.
Every credential has an agreed position j. Partition on it, i = ⌊j/k⌋, at no coordination cost.
3/ Then counting replaces attribution. k tags in a partition => all k credentials spent, without knowing which made which. Delete both lists, zero residual.
Works for zk-SNARK and CLSAG backbones alike, no new assumptions.
This is a joined work with @PrestonVanderV @pedrorechez and Zhipeng Wang
Duc V. Le retweeted
.@circle set out to put dollars on the internet and make them work everywhere. Now @jerallaire thinks the main users of that infrastructure could be machines.
He joins @cdixon to talk about what comes after digital dollars, and what the agent economy needs when software can earn, spend, and hire.
Highlights:
1:58 Circle’s original vision for programmable money
7:19 Why institutions started taking stablecoins seriously
12:04 Learnings from building a global financial network
15:03 What gives a stablecoin its moat?
26:28 Crypto’s growing adoption in the shadow of AI
28:11 From making existing things better to creating entirely new markets
32:59 The weirdest experiments being built on Arc
37:02 What happens when AI hires AI
38:04 Why the next big thing starts out looking like a toy
39:05 Can blockchains counter AI’s centralizing tendencies?
44:01 What the machine economy could look like in five years
Duc V. Le retweeted
1/ Today we’re releasing Aerie: fast post-quantum signature aggregation, using lattice-based SNARKs for lattice-based Falcon signatures. On a laptop (CPU only), it aggregates roughly a full Ethereum slot’s worth of signatures (32,768) in 0.65 seconds.
Duc V. Le retweeted
I'm looking for my next research position.
After a PhD with Christian Cachin, I spent the past four years at the Ethereum Foundation designing and analyzing protocols for Ethereum. My interests are now extending to multi-agent AI as well: what can we still guarantee when some agents fail or misbehave?
Open to research and faculty roles in distributed systems, blockchain protocols, and reliable, secure multi-agent systems. I'm based in Seoul and looking in Seoul, Hong Kong and Singapore, and I'm open to remote roles or the right opportunity in London.
If your group works on these problems, or you know of a role that might fit, I'd be glad to hear from you! DMs open.
Duc V. Le retweeted
AI is getting better at solving hard math problems, especially given recent drops. But what does that mean for the cryptography that protects everything from private messages to healthcare to banking and crypto wallets?
Stanford cryptographer Dan Boneh, with @eddylazzarin, @SuccinctJT, @liz_harkavy, and @smc90, makes sense of the warnings sounded recently. They cover:
- what’s fact vs. assumption
- what’s new and what risks cryptographers have long prepared for
- where quantum computing doesn't (and does) fit
- opportunities for builders
Highlights:
00:37 – What sparked the most recent debate about AI and cryptography
01:35 – Signal vs. noise: What the math does and doesn’t tell us
05:29 – Why cryptography depends on open problems
09:25 – Why the stakes go beyond blockchains
10:43 – How AI could strengthen confidence in cryptography
13:50 – Considerations for users and institutions
20:10 – Where AI and quantum threats overlap
22:32 – Building systems that can change cryptographic methods
23:59 – Hash-based versus lattice-based cryptography
33:07 – Opportunities for builders, and better security tools
38:14 – What comes next if there were a cryptographic break?
