Mitchell Hsing was a grad student at MIT, still running chip equipment from the 1980s that needed a floppy disk to boot.
That frustration became InchFab: a working clean room inside a shipping container.
The trick isn't cheaper equipment for its own sake, but rather smaller wafers. Four-inch wafers shrink the whole fab down to something a startup can own.
The AI chip conversation points at the leading edge: smaller nodes, bigger fabs, government subsidies. Underneath it sits a quieter category, the chips that will never ship in phone volumes, like sensors, specialty silicon and research devices.
A container-size fab opens chipmaking to the builders those fabs were never designed for.
InchFab started with one grad student who got tired of booting a fab from a floppy disk.
Founders...
Investors are watching to see what you do right after they write the check.
Good founders get straight to work. They've got the rocket fuel, now they're launching.
Bad founders come back asking what to do next.
If you’re getting a check, it’s because an investor believes that you already know where you’re going.
They’re giving you fuel, not GPS directions.
Also... too much celebration is a red flag.
Pop some champagne, sure or in my case sparking apple juice. But then get back to building.
Mitchell Hsing was a grad student at MIT, still running chip equipment from the 1980s that needed a floppy disk to boot.
That frustration became InchFab: a working clean room inside a shipping container.
The trick isn't cheaper equipment for its own sake, but rather smaller wafers. Four-inch wafers shrink the whole fab down to something a startup can own.
The AI chip conversation points at the leading edge: smaller nodes, bigger fabs, government subsidies. Underneath it sits a quieter category, the chips that will never ship in phone volumes, like sensors, specialty silicon and research devices.
A container-size fab opens chipmaking to the builders those fabs were never designed for.
InchFab started with one grad student who got tired of booting a fab from a floppy disk.
Austin is now the heart of autonomy. Got Chinese food delivered fully autonomously for the first time via @goautolane. Stays warm. Super reliable. Super fast. No need to tip. And the food was awesome. When will other cities catch up…
Reuters reports Kalshi is in advanced talks to raise $1B at a $40B valuation, up from $22B in May.
Step back and look at what that number says about the category. Prediction markets were a regulatory punchline a few years ago. Too fringe for serious capital, dismissed as gambling with extra steps. Now the fastest-moving investors in the world are competing to fund them, and valuations are doubling in months.
We backed @Polymarket when that dismissal was the consensus. It has since reached a $9B valuation with validation from ICE, the owner of the New York Stock Exchange. The thesis was that markets aggregate truth better than pundits, and the internet would eventually want a price on everything.
When a category this young starts attracting capital this fast, it isn't a bubble forming. It's the world catching up to what the early builders already knew: prediction markets are financial infrastructure, not a novelty.
The ideas that sound too weird to fund become the categories everyone races into eventually.
We led a strategic round in Xverse, backing its push to become a self-custodial Bitcoin neobank.
Buying Bitcoin in the mid-2010s was not necessarily hard because of lack of conviction. It was hard knowing that holding your own keys meant giving up everything a bank does for you.
Xverse closes that tradeoff: a self-custodial vault with Earn, Cash and Borrow rails, so Bitcoin holders can earn yield, spend, and borrow without ever handing over their keys. Since launching in 2021 the platform has passed two million downloads.
Bitcoin DeFi hit record total value locked in 2025 while exchange balances fell to record lows in 2026.
People want their Bitcoin working without a middleman holding it.
Congratulations to Ken Liao and the @Xverse team. 🤝
This AI agent joined 30+ Discord servers overnight and sent 500+ messages. Live counter next to its name: “Humans who noticed: 0”
All it was told: get past every CAPTCHA and accept all the terms. It did. No custom exploit, just browsing and clicking.
CAPTCHA still guards most servers until today. It checks whether something can solve a puzzle. BUT, solving puzzles is exactly what these AI agents are built for. Server owners are now stuck: friction at the door loses real members; an open door can’t tell members from agents.
So over the weekend, we shipped the Solo Discord bot - one that doesn’t guard the join button. Members verify on their own phone and get a real “Human” role. You put that role on the channels that matter (chat, giveaways, support).
Check is: one living person, counted once, AI agents can’t fake it. FHE encryption runs on the phone; no ID needed, no raw image stored.
Shipping the bot today. Free for servers now. SDK if you want the same check in your own app. Try it now and tell us where it gets annoying. We’re reading every reply.
We're proud to share today that Xverse has closed a strategic funding round led by the legendary @DraperVC.
Both @trustmachinesco and @DraperDragon also joined in the round.
Proud of the @AiSeerco and @facticityai team. Their citation harness just hit 98% on a new legal hallucination benchmark—matching the most expensive frontier models at a fraction of the cost. This is how AI becomes usable in court, not just in demos. Generate first. Verify before filing.
Every machine waits for a signal. When it breaks, everything goes blind.
We started Offline Protocol by keeping people connected when networks failed. Today, that same technology runs on hundreds of thousands of devices worldwide.
Now we're taking it further: resilient coordination for physical systems. Machines, people and AI agents that keep working together over whatever path remains.
Where we are today:
→ 350,000+ devices across 140+ countries
→ Nearly 20,000 new accounts in August alone
→ 11,500+ local networks active this month
→ 60,000+ SDK downloads
→ 440,000+ location proofs received in the last 30 days
→ Zero paid acquisition
Earlier this year, we closed a new round led by @DraperVC to turn our consumer scale into a rapidly growing enterprise business.
Over the past month, we've begun working with partners including a global wind-energy manufacturer, an agricultural IoT provider, a fleet telematics platform, a physical AI operating system, and a data streaming platform, alongside teams in aerospace and satellite systems, embedded silicon and consumer wearables. Our pipeline also spans maritime and industrial monitoring, equipment custody, secure device identity, airline catering and restaurant operations.
Wherever machines operate, Offline Protocol coordinates.
Learn more at https://t.co/otJPfxCwrs
The internet has no idea who anyone actually is. For thirty years that was an inconvenience. With AI bots now rivaling human activity online, it's the defining infrastructure problem of the decade.
Aidan McCarty (@aidan_mccarty_) saw it in 2016, when the civic platform he built with his brother Liam (@LiamHaleMcCarty) got flooded with fake and stolen identities. In 2020 they started Verified to fix it, years before the rest of the market woke up.
Aidan joined Brian E. Thomas (@DivergentCIO) on the Digital Executive podcast (@CoruzantPod) to talk through the build. What Verified does: from a phone number and first name, it confirms a real human and autofills verified data in a median of under 10 seconds. No captchas, no document uploads, no losing legitimate customers to the friction meant to catch fakes.
253 million Americans instantly verifiable, 4x growth in the past year, 31% more users converting, and millions of verifications annually across financial services and healthcare.
The founders who start fixing a problem before the world agrees it exists are the ones holding the infrastructure when it does.
Our portfolio company @VincereBio just received a $2.7 million SBIR Phase II award from the NIH's NIDDK to advance its USP30 inhibitor through IND-enabling studies for acute kidney injury.
AKI affects over 13 million people worldwide each year, and no approved therapy directly modifies the disease. Vincere's approach clears damaged mitochondria from injured kidney tissue so it can recover.
This is Vincere's second organ from the same platform: their lead program, backed by a $5 million award from The Michael J. Fox Foundation, is advancing toward the clinic for Parkinson's. Over $8 million in non-dilutive funding from the NIH and MJFF now validates the science.
Congratulations to @Dr_Behrouz, @andy_d_lee, and the Vincere team. 🤝
Bitcoin maximalists will hate this, but leaving Bitcoin untouched is the riskiest thing you can do to it.
The instinct to protect the base layer makes sense. But a store of value that does nothing else cedes every other function of the financial system to chains that were only ever experimentation.
NFTs, tokens, DeFi, all of it was a testing ground. Ordinals instead of NFTs. Runes and BRC20 instead of ERC20s. The experiments proved the demand. The work now is bringing those functions home to the strongest asset.
Bitcoin either becomes a true programmable blockchain capable of replacing the financial system, or it stays a vault while the rest of finance gets rebuilt somewhere less secure.
AI will take everyone’s jobs.
... until they invent new ones.
AI alone probably won’t take your job.
But someone using AI might. So you should lean in to AI.
People react to new technologies in three ways:
1. The Resisters: "AI is ruining everything!"
2. The Ignorers: “I don’t understand AI.”
3. The Adopters: "How can I use this to get better?"
It's not AI vs. humans. It's humans with AI vs. humans without AI.
We’re not cutting our team because of AI.
We’re GROWING it.
Our team with AI can accomplish 10x more than our team without it.
Businesses without websites died in the 90s.
Same thing is happening now.
If you're not experimenting with AI in your industry, someone else is.
And they're going to serve your customers better.
Faster.
Cheaper.
The buggy whip makers of the 1800s became car mechanics.
They pivoted or perished.
Which camp are you in?
We published an article on how we built our own AI infrastructure at Draper to automate every routine task and give our team more time to do the human work in VC.
Two years ago, our firm ran on eight Excel sheets.
Today, one auto-updated base runs everything: 400+ portfolio companies, 15,000+ contacts, 1,900+ meeting notes, searchable in plain English from a phone.
We published the full build log, the 5 decisions we made before building anything matter more than any tool:
• Decide where the data lives first. One base, one record per company. Everything else feeds it or builds on it.
• Build on infrastructure you already own. Airtable, Zapier, Claude. No new vendors. Fixing something means editing a prompt, not shipping a deploy.
• Customize the judgment. An off-the-shelf lead scorer doesn't know your thesis or your passes.
• When the system isn't sure, it stops and asks a human. No guessing where the data has to be right.
• Document every tool like a shipped product, so the next person picks it up in five minutes.
The biggest payoff is the hours it returns to the work only people can do: sitting with a founder, getting to know them deeply, and making the call.
Read the whole blog on how we built Draper’s AI stack in the comments.
The USD.infra Vault Is Now Open 🔓
The Internet runs on digital infrastructure. Now with DAWN, that infrastructure is financeable on-chain.
Introducing the USD.infra Vault: A new way to finance digital infrastructure powering AI compute and connectivity.
Congrats to Alsa @AisaOneHQ team! As Alsa's early investor, we are excited to see the progress. As GTM become truly agentic by default, Agents can now do the work, not just surface insights, while Alsa provides the tools and data needed to research markets, track competitors, uncover search gaps, understand customer conversations, and identify prospects.
Your @dify_ai agent just got a GTM research team.
- Similarweb + Ahrefs track competitors
- Semrush + DataForSEO find search gaps
- X + Reddit read the room
- Apollo finds prospects
AIsa Go-to-Market plugin is live in the Dify Marketplace.
Put it to work 👇