Our latest report on the State of Tokenization is out today! Last time, we asked how much of today's tokenized market is stuck in the "newspaper-on-a-website" phase. This time, we asked what comes after tokenization? How do tokenized assets become tradable markets with broad investor participation? Our Q2 research covers 671 tokenized assets ($330B+) with amazing work by @sui414 @0xallyzach. We updated our Tokenization Progress Index (TPI) across issuance & redemption, transferability, and composability. This edition adds new dimensions: onchain trading and ownership data. These help depict the current state of liquidity, trading volumes, and overall participation in tokenized assets. We also cover key Q3 developments, including @RobinhoodApp Chain's launch and the Clarity Act. Takeaway #1#: Open assets are winning the battle for liquidity. Among tokenized assets of $10M+ value, the open / freely transferable ones made up 41% in value, but processed 99.8% of spot trading. Takeaway #2#: Equity demand is arriving mostly through derivatives, mainly perpetuals. June equity perp volume hit $67.8B, ~16x the $4.2B of tokenized-equity spot volume. Takeaway #3#: Liquidity means something different for each product. Yield funds need dependable redemption. Tokenized stocks need deep liquidity. Collateral needs reliable valuation & liquidation. Despite the CLARITY Act's delay, the momentum behind tokenization can't be stopped. The Great Onchain Migration marches forward, from putting assets onchain to building new markets around them. Here's the full report link:

We've accepted that LLMs will be ubiquitous. What comes next is direct control and access: an LLM in your pocket. Today's accessible hardware is still at early-iPhone levels. The goal is to deliver modern capability so teams can run powerful open-source models without needing multi-GPU setups that cost hundreds of thousands of dollars. @b3labs's rent-to-own model changes the economics: lower the initial outlay, own the machine in two years, and avoid the burden of hosting it yourself. Secure, managed facilities handle the infrastructure. For long-term AI workloads, the math favors ownership over renting. @darylX24 and @viktoriya0x (B3) and @yorkerhodes (Microsoft/NYU) join Stateful, hosted by @FranklinBi, to discuss why the next wave of AI builders is buying instead of renting:

"what if you could OWN (not rent) your own commercial / workstation grade GPUs (H200s, B300s) on leverage, while still earning from these productive assets?" a simple concept - but what if it scales? we'll move from the age of "supercomputers in your pocket" to "supercompute in your backyard"

i think i found it, @b3labs just launched B3IQ basically you own the NVIDIA GPU, they host it for you, and when you're not using it your compute can earn for you so apparently my GPU can have a side hustle now lol

an NYU grad student in @yorkerhodes's lab was researching human trafficking & war-zone evacuations. the frontier lab's safety filters flagged the work and banned her account, sealing off the work and data permanently. that could happen to your business tomorrow. If you rent your intelligence, someone else's priorities & roadmap sits between you and your work. if you don't own your compute, you don't own your destiny. @b3labs has built B3IQ for exactly this. on this week's Stateful episode: we talk through all of it with co-founders @darylX24 & @viktoriya0x & one of their early customers @yorkerhodes from @Microsoft and @NYUniversity. ownership is the endgame of the AI buildout. with B3IQ, you can become an owner today.

The smartphone moment for compute. Smartphones took over the world when carriers changed how you pay. Sign a 2-year plan, pay monthly, and the phone is yours at the end. Today, @b3labs opens up the rent-to-own model for AI with the launch of B3IQ: a new model for owning, running, and monetizing compute. The workhorse machine of this cycle is the 8x H200 node. If you rent these from a neocloud, two years of rent can be equal to the full sticker price... except you own nothing at the end. B3IQ sells you the box for ~30% down. US-assembled, hosted & managed in their 27,000 sq ft Oregon facility, with title in your name and compute in your control. But unlike the iPhone, these assets can pay for themselves. Idle capacity rents out through the B3IQ offtake marketplace, paying down the machine's outstanding balance. Renting compute means someone else sets your pricing, your policy & your priority in the queue. It's the ultimate platform risk for the next generation of businesses. Ask the companies going AI-native across Silicon Valley. Nearly every single one is bringing compute in-house to run & train open-source models. Or ask the researchers at Stanford, NYU, Dartmouth & Univ. of Hawaii who use B3IQ today to accelerate their research on cancer, human trafficking, and war zone evacuations. They need B3's predictable compute costs and data privacy that hyperscalers can't provide. Proud to be on this journey w/ @darylX24, @viktoriya0x & @seangeng since leading @b3labs seed round. The team has been cooking ever since. Now it's time to make compute own-able and accessible to the world. If you believe in the future of AI, own the machines that power it. If you are building with AI, own your destiny and take control of your compute.

Open-source AI is catching up to proprietary frontier models, but the hardware needed to run them are gatekept. @b3labs is launching B3IQ so anyone can own their own intelligence. @darylX24 and @viktoriya0x (B3) and @yorkerhodes (Microsoft/NYU) join Stateful, hosted by @FranklinBi In this episode, they discuss why owning your compute is the next unlock for open source AI: - Frontier labs have locked up hyperscaler capacity through 2030. Everyone else is out of line. - NYU students researching human trafficking and war zone evacuations get account-banned by closed models - Two years of renting an H200 node costs the same as buying it. Rent-to-own drops the entry cost to 30%. - Kimi K3 matches frontier performance on some coding tasks. You just need two H200 nodes to run it. - Idle compute gets matched with offtake, so the machine pays for itself 05:28 Content Filters and Token Limits at NYU 11:02 The Hyperscaler Long Tail Problem 16:16 Why Frontier Labs Locked In Their Model Gains 18:19 How the B3IQ Financing Model Works 20:10 The iPhone Moment for Compute 29:27 Why Own the Box When Neo Clouds Exist? 40:24 The Five-Year Vision

Introducing B3IQ. A new model to own, run and monetize compute. Already being used by faculty, researchers, and student teams at Stanford, NYU, Penn, UChicago, Dartmouth, Waterloo, and the University of Hawaiʻi.

Proof-of-human is becoming the scarcest resource on the internet. Every advance in AI agents is equally an advance in bots indistinguishable from people. When most traffic is synthetic, verifying a unique human becomes base-layer infrastructure. That's the thesis behind @world_id, and why we're excited to continue partnering with World.

PERSONAL UPDATE: I’m joining @PanteraCapital as a Partner on the investment team. I'll be backing founders from Pre-Seed through Series B working on the frontier of finance and the technology that powers it.  Let me share what that means, and why Pantera

Hyperliquid and Polymarket dominate in the West. In Asia, there is no equivalent. TurboFlow is building it. @tonytfxyz (@TurboFlow_xyz) joins Stateful, hosted by @masonnystrom. In this episode, they discuss why Asia has been missing a world-class trading platform and how TurboFlow is building it: - Asian traders operate in communities, not alone. That changes everything about product design. - Prediction markets beat perps for retail: click yes or no on anything. No margin knowledge needed. - $9 trillion in perps volume in 2025. Still only 75K to 200K daily active traders on-chain. - Bear markets are for building. Competitors go to sleep. You lay the foundation. - 18-month goal: dominate Asian prediction markets and expand to 50 assets including RWAs and FX. 00:00 Why Asia Has No Hyperliquid or Polymarket 01:56 The Perps Market Today: $9 Trillion, Only 200K Daily Traders 06:07 Why Prediction Markets Beat Perps for Retail 11:18 What Actually Matters When Building a Trading Platform 14:45 Lessons from Co-Founding Amber Group 18:46 Quick Fire: Bear Market Timing, 18-Month Roadmap

one of the clearest patterns in polymarket’s short-term btc markets shows up off-platform. across ~11.8k polymarket btc up/down cycles, binance spot volume jumped into the close when the polymarket contract was still trading like either side could win. when the contract was already priced away from 50/50, the burst mostly vanished. 1/8

Had a great time chatting with Kili on her new podcast! I’ve known her for what feels like decades in crypto time and I’m super excited to see her launch and grow @wall_collective. Give it a listen! 👇🏼

PSA: my impersonators are out in force again, so please remember that the real me: - doesn’t have $20m of Filecoin to sell you - doesn’t need gift cards - greets you with proper honorifics like “your lordship” or “your grace” or even “blessings upon thee” if we cool like that

Excited to welcome @jshinny12 to @PanteraCapital! What stands out from my first chat with Joey is his obsessive curiosity and truth-seeking, independent thinking - core ingredients of a great investor and partner to founders. Joey's a builder with technical chops and software eng background that stretch across DeFi, MEV, and ZK research. He was also a member of @franklin_dao at @Penn - not named after me, dont believe the rumors... nevertheless, Go Quakers! Thrilled to have Joey on board and excited to work together to partner with a new generation of legendary founders. In fact, we just wrote our first check together this week - so reach out to us if you're building something great!

crypto is no longer one industry — it's at least 4: 1. stablecoins + payments 2. Bitcoin, crypto asset class 3. tokenization + onchain financial services (defi) 4. blockchain infrastructure they are of course inter-related. but increasingly divergent in context which is part of the mixed vibe right now