←
Ching Tseng925 posts
Ching Tseng avatar

Ching Tseng

@chingtsengtw

Princpal at @AppWorks | Sharing views

◷ Joined November 2013▧ 56 Photos and videos
TYC avatar
⟳ Ching Tseng retweeted
TYC@johnny_tyc·Sep 24

I'll be in SF for Tech Week from Oct 3–18. If you're building or investing in AI, I'd love to grab coffee! A bit about us: At @AppWorks, we've spent the last few years backing vertical AI across Asia and US, including medical AI, manufacturing AI, and AEC. Our LPs include Taiwan's leading manufacturers and financial institutions, along with sovereign capital from across Asia, such as Taiwan's NDF, Korea's KVIC, and Malaysia's Khazanah. Where we can help: • Insight into the Asian market landscape and go-to-market • Introductions to Asian manufacturers • A track record of helping portfolio companies build engineering hubs in Asia If any of this sounds useful, my DMs are open! 📸 Took some of our founders hiking in Yangmingshan, just outside Taipei, and ran into these two cuties along the trail.

Post attachment
Ching Tseng avatar
Ching Tseng@chingtsengtw·Sep 7

If a robot performs the same task every day, why invest so much in pre-training? @SkildAI’s S1 suggests an answer: pre-training can buy deployment speed, not just better performance. In one experiment, a robot began attempting a new task only five minutes after watching a human demonstration—with no fine-tuning or weight updates. In this article, I explore why robotics deployment should be measured by two things: final reliability and the time from demonstration to the robot’s first autonomous attempt. The full article is linked below. I’d love to hear your thoughts. chingtseng.substack.com/p/wh…

Ching Tseng avatar
Ching Tseng@chingtsengtw·Aug 24

The robot data market likes to measure supply in hours but the most valuable company may not be the one that collects the most data. What's the key & How different companies like @scale_AI @encord_team @micro1_ai @MeckaAI and @hbr_pbc are approaching different parts of this emerging value chain. I’d love to hear your perspective. x.com/chingtsengtw/status/20…

Ching Tseng@chingtsengtw Aug 24 Article The Bottleneck in Robotic Data Collection remains scarce, but the harder question is which episodes should enter training and whether they actually improve the robot. The robot data market likes to talk in hours: hours of

Ching Tseng avatar
Ching Tseng@chingtsengtw·Jul 30

We backed @smykjain and @0xfluid team🥳

AppWorks @AppWorks Jul 30 Article Why We Invested: Samyak Jain, Founder of Fluid DeFi already has Aave. So the first question anyone may ask is why build another lending protocol. Yet if digs deeper, they will realize capital fragmentation still creates a huge structural

Paul Frambot 🦋 avatar
⟳ Ching Tseng retweeted
Paul Frambot 🦋@PaulFrambot·Jul 21

Morpho Midnight is live. This is our most ambitious step yet, introducing what onchain finance was missing: giving the users the ability to set their rates and terms. Midnight opens up many possibilities for fintechs, institutions, and credit desks alike, offering a level of customization, predictability and control that has never been possible before onchain. This new protocol brings us closer to building the open credit network for the world, and bringing the $200 trillion global credit market onchain. It enables new kinds of markets, which attract new assets and new distributors. Those attract new curators, and new curators bring in even more loans, creating a flywheel that drive onchain credit forward. Let’s fly 🦋

The missing piece for onchain finance is here Fixed rate, fixed term credit markets Morpho Midnight is live

0:55
jaye | sanctum ☁️ avatar
⟳ Ching Tseng retweeted
jaye | sanctum ☁️@eggpanned·Mar 31

More than $2B of SOL is left idle in LPs, bridges and treasuries. @sanctumso's new product will enable protocols to earn substantive yield on users' SOL deposits without changing how SOL is used within the protocol. Designed to replace wrapped SOL, it is highly optimised and will create a significant revenue stream for you. If you're a protocol interested to learn more, reach out to me. Launching soon!

Bill Hsu avatar
⟳ Ching Tseng retweeted
Bill Hsu@cebillhsu·Jul 9

Is your RWA real collateral, or just a receipt? Anyone can wrap the same underlying, but real difference lies lean across 2 axes: - Distance to on-chain cash - Legal claim level An RWA in the top right (close to cash with a claim that survives stress) beats an identical one sitting bottom left every time. As collateral, this asset will have lower haircut, regulated capital can hold it, and and it will win.

Post attachment

Bill Hsu @cebillhsu Jun 26 Article How many steps is your tokenized asset from cash? Does your asset need to turn into something else before it can be used? Sounds like a simple question, but it's actually the question every clearinghouse risk engine in the world answers every single

TYC avatar
⟳ Ching Tseng retweeted
TYC@johnny_tyc·Jul 5

my two cents re: $VVV 1. What $VVV Actually Is Something @hosseeb said in his video got underweighted in all the noise. He's historically been a loud critic of crypto companies holding equity and tokens simultaneously, because of exactly this kind of incentive misalignment. And yet here he is, in the deal. His justification is worth taking seriously. He's investing in Venice the AI company. $VVV, in his framing, is a computing power utility token — not a governance token. And honestly, that distinction does a lot of the explanatory work here. $UNI, $HYPE — those are fundamentally equity-like instruments. Holding $UNI means you hold a fractional claim on the Uniswap protocol. Buyback-and-burn is semantically a dividend in a different wrapper. When Hyperliquid burns $HYPE, it functions like a company returning capital to shareholders. $VVV is doing something structurally different. The core mechanism is: stake $VVV, mint $DIEM. $DIEM is a perpetual annuity token — hold one $DIEM, earn $1 in computing credits per day, with no expiry date. There's no equity in that. On Venice's balance sheet, $VVV is closer to deferred subscription revenue than to issued stock. The analogy that makes most sense to me is airline miles: a liability Venice has committed to honor in the form of future compute. When Venice does a "buyback," they're redeeming that liability. It's closer to an airline retiring frequent flyer obligations than to Apple buying back shares. And what this whole episode really did was force a realization that was always there: people bought $VVV expecting to participate in Venice's upside as an AI company, but it was never that instrument. 2. Why the community has been on fire? A large part of that, I think, comes down to context. The past year has not been kind to token holders. Across dozens of protocols, retail investors watched teams, VCs, and insiders extract value through token unlocks, treasury distributions, and equity rounds they had zero visibility into. The rage is accumulated. This deal brought the structural reality into the open, and people snapped. And the structural reality is: in 99% of crypto deals, token holders have no formal protections. Token conveys no ownership, no voting rights on material corporate decisions, no priority claim on assets. Token holders sit at the junior-most tranche of the capital stack, and equity holders sit above them. What this deal did, more than anything, is make that visible in a way that couldn't be ignored. Beyond that, there's a detail worth naming that I don't think got enough attention. Those 8-year token warrants are essentially free call options. In most primary crypto investments, token warrants are opt-in instruments acquired for nominal cost — sometimes $500, sometimes $1,000, sometimes nothing. Dragonfly invested in Venice the company and got the token upside (with both vested tokens and a call option) basically for free on top of it. To many they went in wanting equity and walked out with more than they appeared to be buying. Whether that's fair is a separate question, but it's part of why the whole thing looks the way it looks. 3. The Buyback Math Is Hard to Make Work Venice has roughly $70M in ARR right now. Yeah, that can be true, and it's impressive. But math still gets uncomfortable. The buyback mechanism is funded only by new subscriber revenue — every new subscriber, depending on their tier, routes a portion of their subscription fee into the buyback pool. Current annual buyback volume sits around $2–2.5M per year. Against a market cap of roughly $650M, that's a buyback multiple of around 220x. Most equity investors start squinting when buyback yields fall below 2–3%. At 220x, the implied yield is under 0.4%. That's a tough number to hold. One can argue, fairly, that Venice is growing fast and the buyback volume scales with subscriber growth. The team could also launch new products, new initiatives that change the math. Both are real possibilities. But because the buyback mechanism is specifically tied to new subscriber revenue, you're ultimately betting that subscriber growth continues at an aggressive rate, more or less indefinitely. And even then, the question is how much of any new value Venice creates flows to $VVV holders versus to the equity holders sitting above them in the stack. 4. Market is pricing it So you have a utility token priced at 220x its annual buyback volume, in a structure where a senior tranche now exists and has every incentive to capture the equity upside of the business. The mechanics of how $VVV works are interesting. The buyback math is what it is. But the question underneath all of it is simpler: when you hold the junior tranche, how much can you realistically expect to benefit from a company's growth — especially once someone senior shows up? That's the question this whole dispute was really about. And it's one the crypto market is pricing right now.

On tokens vs equity, Venice, and $VVV

8:16