There was a time when a computer filled a whole room, and you needed permission to touch it. Only a few ever did. Then the machines got smaller, and the doors opened a little. Not for everyone, though. In India, the first computer for many people was an internet cafe: a small hot room, a screen that was not yours, and a clock ticking over every hour you paid for.
The computer moved onto your own desk. Then into your pocket. The first mobile phones were costly and rare, a thing for the few. A generation later, everyone owned one, and no one thought about it.
Each time, the same thing happened. The technology got cheaper, came closer, and became personal. And each time, far more people got to use it.
AI today stands where the internet and the personal computer once stood. The best models live far away, in data centres. You rent them by the token. Every question starts the meter.
Look at what sits behind that meter. Energy feeds chips. Chips run inside data centres. Data centres serve models. Models power the apps you use. Every layer adds a cost, and today you pay for all of them.
“Every token produced is the result of electrons moving, heat being managed and energy being converted into computation. There is no abstraction layer beneath this.”
Jensen Huang, NVIDIA, AI Is a Five-Layer Cake
That meter is priced for the few. Even in the United States, only about 3% of households pay for AI. Worldwide, fewer than one in five working-age adults use it at all. A monthly subscription is out of reach for most people. And many do not yet know what AI could do for them, so they never ask.
Picture a farmer deciding when to sell. A student who cannot pay for a tutor. A shopkeeper doing the books at night. A patient trying to understand a report. Someone who wants AI to answer in their own language. They live in Lagos, Lucknow, and Lima. AI could change their days. Today most of them cannot afford it, and many have never been shown what it can do.
We build from India, but the problem belongs to the whole world. So does the answer.
Some models will always live in big data centres, for the hardest problems. Most of daily life does not need them. We will find a way to bring that intelligence to the phone in your pocket.
So we work on the layers one at a time, starting from the top: applications, models, and infrastructure. Less waste, more done, and your data stays with you. Then we go deeper, to chips and energy.
Think of the best help you have had. A friend who listens. A mentor who remembers. A colleague who catches what you missed. We are building that for everyone. AI cheap enough to use all day. Private enough to trust.
AI should be cheaper, faster, and yours.
see what we are building →