We sell three things into one enormous market. AI work, run on our subscribers' phones and sold to businesses (YOnC, the Yonder Neural Cooperative). Phone plans and travel data in 200+ destinations, an AI assistant, entertainment, and rewards that pay (Global YO, YO Mobile NA). And, once the acquisition we are negotiating closes, the wholesale data other operators buy (YOx).
The same customers, the same partners, and the same platform feed all three. The numbers below are for YOnC, by far the largest of the three and the one the money raised now goes to build.
*Market size is our own estimate, built on Silicon Data's compute price index and published demand figures: about $210 billion a year of AI work today, rising toward $660 billion over the next decade as fleets refresh. The network's plan draws on a small share of the roughly 4.2 billion AI-capable smartphones in use. Sources: GSMA Intelligence; Counterpoint Research; IDC (International Data Corporation) Worldwide Quarterly Mobile Phone Tracker; the International Energy Agency; Silicon Data; published cloud pricing, September 2026.
Each business earns on its own. The wholesale data YOx will sell to other operators once the acquisition closes. The mobile service we sell to travelers and households. The AI work we sell to businesses. Each one makes the next cheaper to run or harder for a competitor to copy. That is why this beats being a one-product company.
YOx · the wholesale layer · acquisition in negotiation
The network plumbing that supplies data to every YO consumer business and to the operator partners. YOx is formed from an existing wholesale operator that we are in negotiations to acquire. It supplies the YO brands today and comes under House of YO on closing. It earns on its own, and everything else here sits on top of it.
Global YO · YO Mobile · RoamX · operator partners
The apps people actually use, and the front doors that put the network in their hands. Three YO brands are live today: Global YO in 200+ destinations, YO Mobile in Mexico with the United States network in test, and the standalone YOnC earning app. RoamX, the Jazz-branded travel product for Pakistan, is live on the same platform, without YOnC for now. The first YOnC operator market opens this year, with YOnC inside the operator's own application. Everything runs on the same YO platform, and every app is a way into the YOnC network.
YOnC runtime · standalone app + in-app module
A network that puts ordinary smartphones to work running AI while they sit idle, and pays their owners. It needs the other two layers: Connectivity moves the data, and Distribution puts the phones in people's hands. The computing power is already out there. YO connects it and sells it. YOnC is the newest product in the House of YO portfolio, built on what the other two layers already provide, and it is what the money raised now goes to build. The YOnC app is for people who want to join the network without buying a travel or mobile plan.
YOnC, the Yonder Neural Cooperative, is a network that runs AI on smartphones already in people's hands. The same phones people use for travel data, calls, and entertainment take on paid AI work while they sit idle, and the owner gets paid for it. The owner also gets free AI on the device.
See exactly this on your own device, in the YOnC Bench closed beta
A phone that earns its keep puts both modern AI and a new line of household income within reach of people who could never afford a flagship phone and a subscription. YO already built and runs the network infrastructure, the apps, and the rewards system that tie it together. The compute layer is what the money raised now pays for.
↻A closed loop, not a funnel. Every turn builds on the last. Money in, more subscribers, more AI work, more paid back to owners, and around again.
Free AI for the user, income for the owner. The phone runs a capable assistant at no cost and earns real money while it sits idle. That puts modern AI within reach of people who could never pay for a flagship phone and a monthly subscription. Four steps, each funding the next. The subscriber stops being just a customer and becomes a paid part of the business.
Two columns of facts. The left side is what building AI capacity runs into today: physical, slow, expensive. The right side is what changed in the last twelve months. Read across each row. The right card answers the left one.
Three physical limits no amount of capital can compress.
Finding the site, getting the permits, ordering the transformers, connecting the power, and switching it all on takes a big AI data center about five years, before it does a single piece of work. Money cannot shorten that calendar. A lab that wants capacity in 2027 needs it under contract today, and the only capacity it can contract today already exists somewhere.
→ Data Center Knowledge · 5 to 7 year grid waitsIn the US, the waiting list to connect a new data center to the power grid stretches past 2030. Microsoft restarted the Three Mile Island nuclear plant in 2024 just to power AI. Whole projects are being moved to wherever power is still available. Transformers are on multi-year backorder, and permits take years. None of this can be fixed by spending more.
→ CNBC · Constellation restarts Three Mile Island for MicrosoftThe big cloud companies have more than tripled what they spend on AI since 2023, and plan to spend more in 2026. Even so, the top AI labs say what holds them back is capacity to run their models for customers, not to train them. The money is there. The things it can buy are not. A network that adds capacity without sites, buildings, or transformers gets around the problem entirely.
→ Yahoo Finance · hyperscaler 2026 AI capex tops $700BThree structural changes the AI economy just absorbed.
Silicon Data publishes a daily price index for AI computing power, live today. CME Group lists futures on that index from 5 October 2026, pending regulatory review. Once they trade, a buyer of the network's output has a public benchmark to value the position against, which answers a question professional investors always ask. Compute is on its way to being priced and traded like any other commodity.
→ CME Group · compute futures listing 5 October 2026SPAN's XFRA program puts NVIDIA chips in ordinary homes to run AI, with NVIDIA and PulteGroup as launch partners. It proves real AI work does not need to happen in a data center. YOnC runs the same play on the phones already in millions of pockets. No buildings, no permits, no new power lines.
→ CNBC · NVIDIA, Pulte & Span put compute in homesAndroid already lets an app do paid work in the background, so YOnC can ship on Android phones now, and Qualcomm's tools give third parties direct access to the AI chip in most Android flagships with no gate. A phone that earns money for its owner is a feature any maker can sell. iOS is tighter. Apple limits background work, so the iPhone path runs through a background-policy review we have filed, and we treat it as upside rather than the plan.
This case does not assume Apple or Google will welcome us. Android already permits the background work today. The iOS path is under review. John Ternus became Apple's CEO on 1 September 2026. Nothing here depends on it.
→ Apple Newsroom · Cook to Ternus transition, effective 1 Sep 2026Two instruments pay for this build. Priced equity into the company, and a forward sale of the network's own output: a buyer pays today, at a discount to the price of the compute on the day the contract is placed, for capacity delivered later, and takes delivery in compute. The company backstops delivery.
The instrument is deliberately not novel. It is the structure mining, energy and cloud already use to fund capacity before it exists. Terms are shared under agreement with qualified investors. Nothing on this page is an offer to sell or a solicitation of an offer to buy any security.
Pre-purchased output is delivered as verified work over the term, and anything undelivered extends. The company backstops delivery with third-party capacity.
Silicon Data publishes a daily compute price, and CME Group lists futures on it from 5 October 2026, so a buyer can value the position against a public number. That answers a question big buyers always ask.
Prototype running now. First operator market this year. 100,000 devices in the fourth quarter. One million in 2027. Each step has to be met before the network grows to the next.
The forward sale assembles three structures investment banks already underwrite at scale.
Wheaton Precious Metals and Franco-Nevada built their businesses pre-buying future production at fixed discounts to spot, sometimes with equity attached. Combined $13.6B+ in cumulative upfront purchases.
Long-term anchor power purchase agreements underwrite renewable project finance by giving developers contracted offtake at scale. The forward sale of output mirrors this construct.
AWS, Azure, and Google all sell their capacity in advance at 30 to 60 percent discounts in exchange for multi-year commitments. Ours is tighter still.
Each step has to be met before the next begins. The first is running now.
100-handset prototype running. iOS background-policy review filed. Two patent applications on file, four drafted, a seventh with counsel.
The first operator-distributed market opens in stages. The opt-in rate and what each phone earns become measured fact, not a model.
Sales to business open, and the first in-country contracts begin.
The margins prove out: the network's output grows faster than the cost of running it.
Once the network is producing, businesses and governments buy YOnC computing power the same way they buy it from the cloud today: by the token, in dollars, billed monthly. The difference is that our supply is already sitting in people's pockets. There are no buildings or power plants to pay for, and the output is priced at parity with the low-cost commercial models these buyers already use.
Top AI labs, governments, and large companies in places where AI work has to stay in the country. The same buyers who already purchase from the big clouds, now with a second and very different source of supply.
Buyers pay dollars per million tokens of completed work. The network, not the phone, checks and counts the work before it is billed, so buyers can trust the output while phones keep the cost low.
Output is priced by the token at parity with low-cost commercial models, and there is a public benchmark to mark against: Silicon Data's daily compute price, with CME Group futures listing from 5 October 2026. Supply that has to stay in one country earns a premium where governments require it.
Reaching 100,000 phones in the fourth quarter of 2026 opens enterprise sales, and the first in-country contracts begin. The network grows to one million phones in 2027, when revenue against pre-purchased capacity starts to ramp.
How this connects: the capital raised now finances the build that produces this capacity, and enterprise buyers pay for it as it is delivered, in cash.
None of this is a forecast. Three YO consumer brands are live today, and the first operator-distributed market opens this year, all on the same platform. YOnC ships into each of them as an optional feature, the same way any feature would. What follows is what is already built.
No one has shipped this before. That is the bet. What this team has already shipped is the part nobody else can copy: the operator licenses, the sales channels, the live customer base, and the platform under everything. Two United States patent applications filed in April and May 2026 cover how the work is routed across phones and how it is proven, with four more drafted and a seventh with counsel. This is a team of operators with the reach to make a new kind of income real.
A serial entrepreneur (Urban Box Office, YOnder Music). He built the product and the team to bring information, entertainment and now income to people through the phone already in their pocket. He owns the strategic vision, the investor relationships, and the YO brand.
A product and operations leader in financial technology and emerging markets. At Bank of America and at Wells Fargo she ran delivery across programs of more than 200 people in the United States and India, and advised senior management on what to fund and in what order. At YO she has launched markets and negotiated the operator partnerships.
Building YO Mobile's expansion across Mexico and the United States. She directs the Mexico operations, finance and regulatory compliance, and manages the commercial partnerships, including AT&T and OXXO.
A decade building and launching mobile applications used by millions of people. He runs growth, data and personalization across the YO products, and the analytics and pricing work behind them.
Runs the link between engineering, marketing and design, and decides what gets built first. Owns the schedule that turns what the runtime can do into shipped product, and holds releases to the launch milestones.
Existing capital comes from operators, family offices, and private investors who underwrote the platform that became House of YO.
Bill Anderson, entrepreneur and private investor through his holding company June Street Capital, LLC. Former Executive Chairman and CEO of Anderson Holdings, a family-owned, diversified holding company with subsidiaries in real estate, beverage, and insurance.
Q Prime, manager of Metallica, Muse, Eric Church, and more. One of the most consequential figures in modern music management, and an early investor in YMM.
Nadhim Zahawi, co-founder of YouGov and former UK Member of Parliament who served in several ministerial positions and as Chairman of the Conservative Party under Prime Ministers May, Johnson, Truss, and Sunak.
Guomin Xie, founder and former Co-President of Tencent Music. Invests in YMM through GMC Assets Corporation (BVI), his family office.
Strategic-stage investor in connectivity, distribution, and consumer mobile platforms.
The company is raising capital to build the network. The fastest way to understand what it is building is to run it on your own phone.
YOnC Bench is in closed beta. It is a free Android app that runs a real AI model on your phone. No internet, no cloud, nothing leaves the device. It measures what your phone can do, then shows what that work would be worth on the network.
The app runs an open AI model on your device. Switch the internet off and watch it answer. Nothing is sent to a server.
A sixty second test measures how fast your phone runs the model. Then it shows what this exact device would earn on the network. Not a slide. Your hardware.
No data leaves the device. That keeps it private, and it is also why the network can pay people in the first place.
Invitations go out in waves. Keep an eye on your inbox.
