early stage crypto fund accelerating the programmable future
NYC
- Tweets 1,899
- Following 161
- Followers 11,616
- Likes 11,182
Thank you to everyone who joined us and @TACEO_IO for the launch of Merces last night!
And to our speakers @Matteomanzi09, @cameron_nili, @Jessesawa, @JanCamenisch, and Jonathan Mayeur
The privacy layer for onchain finance is officially live
TACEO Merces is live on mainnet.
Real funds, on Monad and World Chain.
Public chains give institutions settlement, interoperability and composability. They also make counterparties and amounts visible to anyone with a block explorer, and the usual response has been to work in a walled-off environment, or to stay off the public rail altogether.
Merces makes privacy a property of the rail itself. Balances and counterparties stay hidden onchain, and no node in the TACEO network can read a balance, because no node ever holds one in readable form.
The state stays shared even while it stays private, which is what makes this more than private payments. A stablecoin transfer may only needs the transaction hidden. A tokenised fund, security or other tokenised assets (RWAs) needs a private register that still enforces eligibility, concentration limits and pro-rata distributions. Rules have to look across every holder at once without revealing what any one of them owns.
Financial logic still runs, and every operation produces a proof, verified onchain, that it executed correctly. Correctness is proved rather than trusted, so it doesn't rest on the integrity of any individual node.
Compliance works the same way.
Screening happens before a transaction executes, and afterwards the institution decides what gets opened to an auditor or a regulator. Specific records, not the whole book.
None of it requires migrating migrating your stack.
Merces works with the chains, token standards and keys already in use, and the same integration works across chains. No new chain, no new asset, no new wallet.
Payments settle on mainnet today. Yield, agentic payments, and issuer controls run on testnets, moving to mainnet through the rest of the year.
@monad and @world_chain_ are the launch networks.
@LayerZero_Core is building private cross-chain transfers with us.
@IntellectEU is integrating Merces into CatalyX for its bank clients.
@OrionFinanceAI is working with us on private vaults.
@ZenithNetwork is working with us on shared-state privacy inside Canton.
2
1
16
1,676
The onchain rails institutions want are finally getting the confidentiality they need
Congrats to the @TACEO_IO team on launching Merces!
The privacy layer for onchain finance is now live.
Meet TACEO Merces
Bank infrastructure, asset managers and cross-chain protocols already integrating, read more: core.taceo.io/articles/merce…
2
17
941
Should LLMs be allowed to write circuits and proofs?
If you have feelings about this, we have an event for you
We're cohosting a @WPReadingClub night with @kobigurk and Abhi Shah on the unlocks and challenges of formal verification
Please RSVP below if you'd like to join us!
3
4
10
766
Whitepaper Reading Club: Formal Verification in Cryptography
luma.com/wprc-nyc-260929-for…
3
152
Archetype retweeted
I couldn’t be more thankful to work with such an industry titan….
Chris is an incredible human and legendary builder. We’re beyond fired up to have him on the team at @octane_security
If you’re keen to work with @dildog and the Octane team, shoot me a dm! We’re hiring
🚀 📈
I am pleased to announce that I’ve accepted the role of Chief Scientist at @octane_security!
I’m excited to be working with @giovignone, @paologentry, and everyone on the Octane team to help build the next generation of their awesome AI-powered security analysis service.
ALT Octane Security
7
7
35
3,030
There's a huge opportunity to connect onchain capital with productive assets in the real world
And as AI infra continues its massive build-out, financing it onchain is a natural fit
@dawninternet is early here
Since 2024, DAWN has raised $40M to date to tokenize AI and connectivity infrastructure.
This has been led by @DragonflyVC, @polychain, @vaneck_us, and more amazing investors.
We’ve spent that time building. Here’s what’s ahead 👇
8
6
71
4,274
Archetype retweeted
Since 2024, DAWN has raised $40M to date to tokenize AI and connectivity infrastructure.
This has been led by @DragonflyVC, @polychain, @vaneck_us, and more amazing investors.
We’ve spent that time building. Here’s what’s ahead 👇
260
237
1,705
719,699
Archetype retweeted
100M+ agentic payments have already settled onchain. The agent economy is valid and operational. So what does a chain built for machine agency actually look like? A map of Ritual, in 6 posts 🧵
72
71
227
24,344
Humans read the web for free
But agents have to pay before they can even decide if a webpage is worth reading
So to help us better understand the costs and friction this creates, Archetype's @oddhash designed a lens that lets us see the web through the eyes of a machine
Stripe’s homepage has 637 KB of HTML for 539 words.
Humans read it for free, but a bot needs $0.26 to read the same words, because it wades 87,400 tokens of markup, if it even gets there.
With this in mind, I built a lens that shows you any webpage the way machines see them
4
1
11
2,452
Archetype retweeted
There are two rates that matter in the macro context of ai models:
1. Frontier expansion rate - The speed and economic significance with which new frontier capabilities are created.
2. Frontier diffusion rate - The speed at which frontier capabilities become available through widely adopted economic substitutes.
The US (1) and China (2) have dominated their respective lanes, and the geopolitical posturing is around defending these lanes.
Chinese models are near-parity on benchmarks, dramatically cheaper per task, and topped US models in OpenRouter token share for the first time in 2026, all through legitimate contributions across quantization, attention architectures, MoE sparsity, and RL.
In my newest report for @Delphi_Digital, I walk through the labs' progress across three efficiency themes and discuss:
- why efficiency gains get competed away as price cuts
- the evolution of licensing and business models over the past year
- why China's grid is a structural advantage
- the distillation fight, escalating sanctions threats, and scenarios for how the US-China AI market splits
- whether Chinese labs can pre-train frontier models on domestic silicon
- implications for investors
Full report below.
3
3
13
778
We were friends & backers of @privy_io a couple pivots before it became the company that powers embedded wallets for 150M accounts across 2.5K apps
@dberenzon joined @segall_max to retrace the journey from finding PMF to building wallets for humans & agents at this year's SBC!
@segall_max, COO of @privy_io (a @stripe company), in conversation with @dberenzon of @archetypevc on secure wallets for humans and agents 👇 piped.video/watch?v=WRpf1hY9…
4
4
28
2,886
“What’s the point of running stuff in an enclave if you can’t verify byte-for-byte everything you built?”
@ZekeMostov of @turnkeyhq explains Turnkey’s Verifiable Cloud and how it makes sensitive AI apps verifiable and easier to build
From our Privacy AI Research Day w/ @monad
2
5
24
5,828
Making the actions of AI systems private and auditable is becoming more difficult, but even more important
@marvin_tong, Co-founder & CEO of @PhalaNetwork, explains the role of confidential compute in increasingly autonomous systems
From our Privacy AI Research Day with @monad
4
2
30
1,429
“The original web3 started as a movement against the big banks. Now we have the big models.”
@KushalBabel, Senior Researcher at @category_xyz, discusses how and why to run AI models locally in a self-sovereign setup
From our Privacy AI Research Day with @monad
3
5
32
3,666