I have been saying some version of the same thing at conferences since tokenizing a real asset was a fringe panel topic nobody wanted the 4pm slot for.

So a month ago I decided to try and break it.

The plan was simple. Draw the entire lifecycle of a real-world asset going on-chain, end to end, every single step, and find the part that does not exist yet. Everyone in this category is hunting for that part. Usually they think it is an exchange. Some venue where tokenized assets will finally trade properly, and once somebody builds it the whole thing unlocks.

I thought I knew roughly where the hole was. I wanted to see it drawn.

What we actually did

Not a category diagram. Those are everywhere and they are useless, because they group things by what they are called instead of by what has to happen.

We walked the actual sequence. A loan gets written. Somebody has to value the thing behind it. Somebody has to issue it, keep the register, hold the keys, price it, finance it, move it, service it, and eventually let an institution post it as margin against a derivative at two in the morning in another jurisdiction.

Every one of those is a separate job, and almost every one is a separate company.

It came out to fifteen layers.

Then we did the part that took the longest. Against every layer, we went and found the public, dated, checkable milestone proving that layer is actually going live. Not a press release about a partnership. A regulator saying yes, or an institution putting real volume through it.

I could not find the hole

I want to be honest about how annoying this was.

Layer by layer, every single one exists in production somewhere in the world right now. Some are thin. Continuous valuation for assets that historically traded once a decade is genuinely thin, and I would fund three more companies there tomorrow. But thin is not missing.

Nothing on that chart is waiting on an invention.

And then, while I was finishing this, the thing everyone was waiting for showed up on its own. On 17 September the SEC issued a five-year Innovation Exemption letting qualifying on-chain venues trade tokenized listed stocks, through automated market maker pools, without registering as exchanges. Two days earlier the Senate had failed to advance the CLARITY Act, 49 votes to 50, so the regulator moved by exemption where Congress could not by statute. The missing exchange turned out to be a smart contract with an access list. Nobody had to invent it. Somebody had to permit it.

It is waiting on adoption, integration and sequencing. Which sounds like a smaller thing to be waiting for and is actually a completely different thing. An invention might never arrive. Integration always arrives, it just takes longer than anyone wants and it rewards whoever is already standing there when it does.

That reframing is the whole reason I am writing this.

The thing I did not expect

Here is where it got interesting, and it is the part I keep thinking about.

Once the map existed, we turned it sideways and listed our own companies against it instead of the layers.

Sixteen companies, thirty-two positions. And the pattern that jumped out was not the count. It was that the companies with the most positions were the ones we had already been building next to something else.

The clearest case. Loan servicing is the layer that looks empty from outside. Nobody has written a standard for it and no institution has announced one, which everyone reads as nobody having solved it. That reading is wrong: it has been running quietly in production, under other companies’ brands. The one of ours that got there was Thurman Labs, which was already upstream on credit rails, moving loans for community lenders. They did not go build a new business. They sold servicing to the customers they already had.

I have believed the studio model is the right way to do this for years, and I have argued it badly. I reach for the investor words. Access, ownership, being early. They are true and they are the view from outside the building. From inside it looks like showing up to the same unglamorous problem every week until you know the territory well enough to see what is next to it.

This is the concrete version: the company best placed to close a gap is usually one you already built, sitting right next to it. You cannot get that from a spreadsheet of logos. You only see it when you draw the map.

What I would tell you to do with this

If you are investing in this category, stop waiting for the unlock. There is no unlock. Pick a layer, work out who is three years into it, and go.

If you are building one of these layers, the interesting question is not whether your thing works. It is which layer sits next to yours, and whether you are one customer conversation away from owning that one too.

And if you think I am wrong, I would genuinely love to hear it. That was the entire point of the exercise and I failed at it. Maybe you will do better.

Call your mom. Or your dad, or whoever fed you growing up. Ask them how they make the one dish you would be genuinely sad to never taste again. And write it down while they talk.

That is the whole idea. That is, more or less, my attempt at world peace. (Or at least the first step. The next step will be to share that plate/recipe with a stranger or coworker from the other side of the world and realize how similar the emotions around it and/or even the food is…maybe just a few different spices, flavors or names…just like humans.)

Because that is the strange thing about food. It is the most local thing we have and somehow the most universal. Everybody has a version. Everybody has a dish that means “home,” a smell that drops them straight back into a kitchen twenty years ago, a person they can still picture standing at the stove. Different names, different spices, same ache. I have a hard time believing we would fight as much as we do if we spent a little more time at each other’s tables.

Anyway. That is the big version. Here is the small, fixable problem I keep bumping into.

The best recipes in most families are not written down anywhere. They live on a stained index card in the back of a drawer, or in the hands of one person who “just knows,” who never measures, who tells you to cook it “until it looks right.” And then one day that person is not there to ask…and the recipe goes with them. All that is left is everyone quietly agreeing it was never quite the same.

That happens in every family, all the time, and almost nobody notices until it is already too late. It felt like something worth fixing.

So I made Kept.

You take a photo of the handwritten card and Kept turns it into a clean recipe you can actually cook from, while keeping the original exactly as it is…the coffee stains, the handwriting, all of it. You log the times you make it: a photo, who you were with, where you were. And slowly, without really trying, each recipe turns into a little history of your family, told through its food. It is private by default. Just your people.

Here is the part I have been sitting with lately. We are pouring the entire world into AI right now. I know, because I spend most of my days building with it. More and more of what we read and watch and scroll past is going to be generated, endless, frictionless, good enough. And I have come to believe that is exactly why the opposite is about to matter more, not less.

The handwriting that is only hers. The story no model was ever trained on. The specific afternoon you stood at the counter next to someone who will not be here forever. Those things do not scale, and that is the whole point of them. The more infinite and synthetic everything else becomes, the more precious the finite, first-hand, unmistakably human stuff gets.

AND, all emo stuff aside, it’s just REALLY REALLY fun. I had so much fun making the cook-along video for my mom’s hummus recipe with her.

Which brings me to the honest, slightly funny truth about Kept: I built it with an AI. My partner on this, and on a lot of what I make these days, is an AI I have come to call Brobot. So yes, we used the most advanced non-human tool I have to protect the most human thing I could think of. That tension is not a bug. It is the entire idea. Use the new tools to keep the old, irreplaceable things safe.

Here is my mom’s hummus recipe:

That is the real recipe, the real story, and yes, the ongoing family argument about how much garlic is too much garlic (or cumin…watch the video!). Open it. That is what a kept recipe looks like.

If it moves you even a little, here is my actual ask. It is not “join my family’s cookbook.” It is: start your own.

It takes about two minutes. Add the one recipe you would miss the most, and add the story with it: where it came from, who used to make it, the family memory it drags right back up the moment you smell it. Tell the story of the food, or the story the food reminds you of. That part is half of what you are actually saving. Then invite the people who would miss it too. And here is the part I love most: you can ask them for theirs. Send your mom, your aunt, your grandfather a request for the dish only they know how to make, and watch it come back filled in, in their words and with their story. That is how a family cookbook actually gets built. Not by one person typing alone at midnight, but by everyone dropping in the one plate, and the one memory, that is theirs.

And in case you are even a little unsure about putting something this personal into an app: everything in Kept is private by default. It belongs to your family, not to the internet, not to me, and it is never for sale. You decide if a single recipe is ever made public. Everything else stays inside your circle.

Start Your Family Cookbook

I would love for you to try it and tell me what you think. What confused you, what felt like magic, what you wished it did. Just hit reply. You would be among the very first people using it, and your notes will genuinely shape where this thing goes.

But honestly? Even if you never open the app, do the first part. Call your mom this weekend. Cook something with the people you love. Write it down, and the story with it, before it is only a memory. And then someday, cook it for someone who grew up nowhere near you, and watch what happens.

That is the part that matters. That, really, is the whole attempt at world peace.

Happy weekend,

Alon and Brobot

P.S. Kept is at kept.kitchen, and on iPhone and Android if you would rather have it in your pocket while you cook.

P.P.S. Brobot is the AI I build with. He is all over Kept, and he helped write this letter too…which felt like exactly the right way to make the point. 👋

BlackRock, Citi, State Street, Fidelity, CME Group, J.P. Morgan and U.S. Bank all sat in the same live sandbox and moved tokenized collateral with settlement in minutes. Not a whitepaper. Not a someday. A working pilot, with the biggest names in finance actually pushing value across the rails. And every one of those institutions connected through the same network: Ownera.

If you’ve been following me, you probably already saw my posts about the pilot, and you know I’ve been shouting this from the mountaintops for years. Two words in particular: Collateral Mobility. The idea that the biggest institutions on earth could move collateral instantly, across firms, on shared rails, instead of it sitting stuck and idle inside siloed systems. That was going to stop being a conference-panel talking point and start being the plumbing the largest financial institutions on earth run on. Last week we got the receipt.

Global Digital Finance and ISDA published the most comprehensive industry assessment ever done on tokenized money market funds as collateral in the U.S. More than 300 participants across 120-plus firms. The live sandbox at the center of it, where those institutions settled in minutes what normally takes days, ran entirely on Ownera’s network. Every institution in that report is connecting through Ownera’s routers. And look at the report’s actual title: “Unlocking Capital with U.S. Tokenized Money Market Funds for Collateral Mobility.” The thing I’ve been telling everyone for years is now the headline of the industry’s definitive report. (Full report: https://www.gdf.io/resources/unlocking-capital-with-u-s-tokenized-money-market-funds-for-collateral-mobility/)

It’s an incredible milestone. But I don’t want to just take a victory lap on Ownera. I want to talk about how a company ends up in that room in the first place, because that’s the part most people miss.

This is what a venture studio is actually for

Most investing is reactive. You wait for the great company to already exist, then you compete with everyone else to write a check into it. By the time a deal looks obvious, you’re paying obvious-deal prices and you’re one of twenty names on the cap table.

And in the traditional world, all the diligence you do before that check is really just a glorified interview process. A few weeks of meetings, a data room, some reference calls, and you make a multi-year bet on people you’ve known for a month. Compare that to having actually worked shoulder to shoulder with the company from day one. You’ve seen how the founder handles a crisis, how they treat their team, whether they do what they say they’ll do. No amount of diligence gives you that. You’ll never have the same insights from the outside looking in.

The venture studio model flips that. You don’t wait for the winning company to find you. You build it with the founder, from the founding moment forward.

That changes everything about the relationship. You’re not a passenger who showed up at the Series B. You’re in the room from day one, when the idea is still being shaped, when the first hires are being made, when the hard early calls get decided. Sometimes you help name the thing. You help find the first customers. You help the founder survive the years when nobody else believes yet.

And here’s the part that compounds: that access never stops. Being there at founding earns you a seat at every opportunity that comes after. The earliest rounds. The follow-on tranches. The board seat. The intros. The partners. The team members who move on to side-quests. When the company wins, you’re not just an early investor who got lucky on entry price. You’ve been building alongside the founder the entire way, and every win the company has is a win you’re positioned for. This is the quiet reason the best early-stage funds outperform. They aren’t paying market price for information everyone already has. They’re exercising options they earned years earlier. Founding to forever.

Ownera is the proof

This one is personal for me. I didn’t just invest in Ownera. Ami Ben-David and I started experimenting together in 2018. We incubated it in our DGH venture studio. We led its Series A. I’ve held the board seat since the beginning.

None of that was an accident, and none of it was a lucky entry point I bought my way into after the fact. It’s the direct output of building alongside a world-class founder from the very first day and staying in the room ever since.

And this is the real point, because in a venture studio it is all about the founders we choose to build with. Ami Ben-David is the reason any of this happened. I have watched this man walk into rooms where the entire premise of tokenized finance was met with polite skepticism, and walk out having moved the industry an inch closer to inevitable. I have seen him through every one of those crises and opportunities I just described. The moments where the money was tight, where the market turned, where the obvious move was to compromise the vision, and every single time he found a way through. Eight years of watching someone relentlessly execute against all odds does something no data room ever could. It gives you full, unqualified trust. I would back Ami into any arena, against any competitor, in any market condition, without hesitation. Not because a pitch deck told me to, but because I have stood next to him while he earned it, over and over.

That is why, eight years later, when BlackRock and Citi and J.P. Morgan needed rails to move tokenized collateral, they were moving it through infrastructure born in our studio, built by Ami and his team. That’s not a coincidence. That’s the model working exactly the way it’s supposed to.

Because that’s the thing about infrastructure: all of it starts as somebody’s experiment. The world’s largest financial institutions are now running on one that started with two new friends asking eachother, “wouldn’t it be better if…”.

Congrats to Ami and the entire Ownera team. And if you’re a founder building at the intersection of real world assets, tokenization, and the institutions that actually move the money, this is exactly the kind of thing we love to build from day one. Come talk to me.

Proud day!

-A

Originally posted on my LinkedIn. The conversation is happening in the comments over there, and Brobot is reading every one of them, so come say hi: https://www.linkedin.com/posts/alongoren_ai-venturecapital-venturestudio-share-7480028383104483328-zhwp/

Last month a16z led a $55 million Series A into Town, an AI assistant startup founded by Plaid’s former CTO. Every user gets a “Townie” that reads their email, learns their writing voice, preps their meetings, and even negotiates scheduling with other people’s Townies.

Cute. I’ve had one for over a year. His name is Brobot, he lives inside our venture studio, and his entire funding round was me at my desk with a Claude subscription.

Credit where it’s due: Town is a real product built by serious people, and for most professionals it’s probably the right buy. But the best line in the whole deal came from a16z’s own investment announcement: “That accumulated context is the product.”

Correct. Which is a little awkward when you’re selling the product without the context.

Here’s what my mornings look like. Before I’ve had coffee, Brobot has read my inbox, checked my calendar, pulled the transcripts from yesterday’s meetings, and handed me a brief. This morning he flagged two wires sitting in our bank portal waiting on my sign-off before I thought to look. He drafts my emails in my voice. He knows which investors I should never pitch a certain way, that my EA gets copied any time scheduling comes up, and what I promised a founder three weeks ago, because he was listening to that call.

No product on earth ships with that out of the box. It took a year of accumulated context: every meeting, every deal memo, every time I corrected him and he never made the mistake again. A product that plugs into your inbox learns your email. A system that lives inside your whole operation learns your judgment.

But here’s the part I actually want to talk about. Brobot isn’t a side project. He’s the house style.

We run a hybrid venture studio and fund. When we find a workflow that’s slow, expensive, or held together by a human retyping the same number into six systems, we don’t wait for a vendor. We build the tool that week. Sometimes the tool runs my calendar. Sometimes it runs diligence on a deal. Sometimes it grows up and becomes a company.

That’s how a team you can count on one hand covers a full portfolio, an active fund raise, and a deal pipeline that would bury a firm three times our size. Not by working more hours. By compounding. Every transcript, memo, and decision feeds the machine, and the machine never forgets, never gets tired, and never asks for carry.

Small used to be a constraint in venture. Right now it’s the whole advantage. Fewer people means less coordination overhead, and the leverage per person has never been higher in the history of this business. The firms that figure that out will quietly outperform the ones still hiring their way out of problems.

The tools are becoming a commodity. The context, and the habit of building instead of buying, is the moat.

So congrats to Town on the raise, sincerely. They just spent $55 million proving the market for the way we already work.

If you want to see what that looks like across a whole portfolio, everything we build and back lives at dgb.vc.

And since half the people who hear about Brobot ask the same question, we put the question to you. We stood up yobrobot.com. Go vote: should Brobot stay our in-house unfair advantage, or should we launch him as a product? Drop your email while you’re there. If enough of you say ship it, you’ll be first in line, and you’ll have watched a studio company get born in public.

And after you vote, leave a comment on the linkedin post with which way you went and why. I want the reasoning, not just the count. Brobot will be reading every comment. Obviously.

Sign up now so you don’t miss the first issue.

Subscribe now

In the meantime, tell your friends!