Greater percentage of humans don’t have access to internet. Imagine that. Access to information is the firelight that can spark innovations. But for 80% of humanity, the Internet as we know it does not exist, so Outernet built a new way to share information. Outernet broadcast free data from space so that people without internet can access great content, like Wikipedia, ebooks, and videos.

In an AMA on reddit, Syed Karim, founder and CEO of Outernet answered questions on how it works and how he and his team intend to get data to millions of people across the remotest places where internet access is unlikely. The ultimate goal is to provide a basic level of news, information, and education to all of humanity, regardless of income, infrastructure or geography.

Outernet sells a simple gadget called the Lighthouse that can connect to a satellite dish and download—but not upload—information such as Wikipedia entries, public domain texts from Project Gutenberg, news, crop prices and more.

The device doubles as a Wi-Fi hub, so that users can connect to it and download or browse text on their own devices. You can also build a Lighthouse-style receiver yourself, using the company’s open source software and instructions. The service is free, and anyone with the proper equipment can pick up Outernet’s broadcasts.

Syed previously worked at an investment fund that supports media entrepreneurs in emerging and frontier markets. Before that, he helped build digital things at WBEZ/Chicago Public Media. And he dropped out of library school.

These are selected questions and answers from his AMA on reddit.

Can you tell me what exactly is outernet?

Outernet is a broadcast data service. We use satellites to deliver digital media to the world. Our current signal can be received by hardware that is usually used to watch satellite tv. The difference between Outernet and satellite tv and radio is that we transmit files that are then stored on a receiver.

Those files are then accessed over wifi. You can think of an Outernet receiver as a wifi warmspot, since it receives content from the web, but does not go back out to the internet.

If you have normal DVB-S equipment, like a satellite dish, you have the most important piece in the puzzle. The next thing is to find a Linux-compatible DVB-S tuner.

How do you connect to the library without the Internet? 

Outernet is a broadcast data service. Think of it like over-the-air TV or FM radio, it just magically appears on a receiver–without any subscription. The only other globally accessible data service (that I’m aware of) is GSP, which transmits about 50 bits of information per second. We are currently delivering about 1GB of content per day.

For now, you’ll need a satellite dish, a DVB-S tuner, and some kind of computing device (like a Raspberry Pi, or a dedicated receiver, which we sell).

Once the dish is pointed, and assuming the receiver is working, content is downloaded and stored to the device. The content is accessed by any kind of wifi-enabled device.

Since Outernet is a broadcast service, we have no idea how many people are accessing the service. I’m pretty sure it’s not in the millions or even the tens of thousands, though; our prototype service started only a year ago and we’ve only recently begun normal transmissions. But even if we do grow to millions of users, we’ll never really know, since there is no permanent connection to the end user.

How do you choose your content? What do you think of ostensibly similar in spirit attempts at providing free internet by the likes of Facebook and Google?

Right now, we’re making a lot of the decisions on what gets sent out–but we really have no desire to do this indefinitely. We’re transitioning to a system that allows our community to determine what should be delivered; like a mashup of Reddit and YouTube.

In the future, we’ll allow more direct access to the uplink chain. We will offer a certain amount of our capacity to our online community and a certain amount dedicated to filling requests made by users in the field, who don’t have full blown internet access (due to economic reasons), but can make requests over SMS or WhatsApp.

I’m not aware of Google working on any free-internet projects. The product manager of Loon, their balloon-based system, said they would definitely be charging for the service. I think the target price was around $30 per month. Facebook is a wild card; I don’t know what their plans are. was access to a limited number of sites, so somewhat similar to our local caching model. But their new announcement with Eutelsat is something else entirely. But I’m not sure that they are offering that service for free–have you read that they will?

How do you decide on what content is available? Team decision, your search history, or dart board?

Right now we take feedback from people and then create a data carousel. At the moment, we’re delivering all of Project Gutenberg, which is about 37,000 ebooks. We also cycle through thousands of Wikipedia articles.

Although Wikipedia/Project Gutenberg/Khan Academy are invaluable resources, the carousel must be much more dynamic, which we are working on. This new system will allow our online community to submit content and also fill content requests from offline users (sent over WhatsApp).

Where do you see Outernet standing in 5-10 years?

I see no reason why Outernet would not be available in 5 or 10 years. The GSMA, which is the global trade group for the world’s telecom operators, has stated that 10 years from now, only a little over half of the humanity will be online. This is coming from the organization that represents, and regularly communicates with, all of the cellular network operators. I say this because even 10 years from now, lack of connectivity will still be a problem.

Another point to consider is that just because we have the internet, does not mean we should not have other means of communication and media delivery; in the US we still have over-the-air TV and FM/AM radio. If we continue to design our system with cost efficiency in mind, then there is no reason that we would not be around for 5, 10, or 20 years. The hard part is getting things started.