Interview with Cliff Schmidt

Interview with Cliff Schmidt – Literacy Bridge

Interviewers: Scott Swigart and Sean Campbell

In this interview we talk with Cliff. In specific, we talk about:

Scott Swigart: If you wouldn’t mind, would you please take a minute to introduce yourself and your project, for the benefit of people who might not be completely familiar with it?

Cliff Schmidt: Sure. I’m the executive director of Literacy Bridge, which is a nonprofit 501(c)(3) charity that’s also a tech startup. We’re building a digital voice player and recorder that is targeted to cost between $5 and $10, depending on volume.

Its purpose is to allow people in the very poorest parts of the world to get access to knowledge, often about crucial things like how to prevent your child from dying of dehydration or how to improve the yield on your crops, or any kind of knowledge that people are after.

The project approaches this problem of illiteracy from two directions. First, part of the problem of getting knowledge is being able to read, so we focus on allowing people to improve their reading skills.

Second, the project also doesn’t wait for someone to learn to read in order to get knowledge, because it is sort of a substitute for the Internet. In places without electricity and without any sort of network infrastructure, this device allows you to get knowledge that’s important to you, spoken in your own language, delivered through device-to-device copying, among other distribution paths.

Scott: Dig into that just a little bit. In addition to the fact that it’s kind of a recorder and player, how does it help people build their literacy skills?

Cliff: It’s meant to be complementary to some sort of existing education program, such as a formal primary school program or an adult-oriented night education class.

What it does is very simple. It starts with anyone reading an existing piece of text, whether it’s a schoolbook, the alphabet written on a chalkboard, or some existing brochure that passes on information. In many cases, it will be a textbook aimed at improving literacy skills.

After someone has read that text, which could be anyone in the community who’s able to read, then the person who’s learning to read can sit down with this device and practice their reading. They can listen to the device read that same text back, jump around from page to page or line to line or word to word, and have vocabulary words defined for them.

It’s a similar process to what you might have if you had a teacher sitting right next to you helping you read, or a parent reading to you. Most parents realize that’s a really critical part of early education, but in some areas of the world, many of the parents are illiterate, so the children don’t have that opportunity. This is a means for them to have a voice reading the book to them, wherever they are.

Scott: Tell us a bit about the technology. You mentioned that it can transmit information directly from device to device.

Cliff: It’s fairly simple, although we’re doing some fairly innovative things in software. The hardware is based on off-the-shelf components, and for the device-to-device copying, we use USB, so just plugging one device into another allows people to share content.

Through this sort of viral distribution of audio content, I think we’ll see that information that’s important to people can travel pretty efficiently through the traditional method that people exchange information, which is face to face. In addition to talking to each other, they can also transfer a copy of the content that one person found valuable.

Scott: You also mentioned that these devices are intended for areas where there isn’t necessarily good power and network infrastructure. How do you address the power requirements of the device?

Cliff: That’s an important question. We start by using whatever is the most accessible and most commonly used source of power. And in the most rural, remote, and poorest areas of the world, that tends to be the large, size D batteries. And these are the batteries that people can find in their local markets.

Since that’s what people use today for their flashlights and radios, we have designed the device to work with the same sorts of batteries so that we don’t have to introduce a new form of power in addition to introducing the device.

Scott: How much battery life do you get off the typical D battery?

Cliff: We need to do some more tests on that to get a good, solid number. There’s a big difference in battery life depending on whether you’re listening through headphones or through the speaker, but it’s going to be somewhere in the range of 15 to 25 hours, I think.

We have a prototype with all the hardware and mechanical stuff is finished. We’re just tweaking some of the software now.

Scott: One of the other projects that we talked to is One Laptop per Child. Talk a little bit about the differences between the two approaches, because it seems like yours is a much more targeted kind of problem and solution. Is there any relationship between the projects?

Cliff: Yeah, there is a relationship between the two projects; we get together every few months and talk about how we might collaborate.

The idea for this project actually started after I had had some meetings with OLPC about possible literacy software that could run on their machines. And it was during one of those meetings, actually, in their offices in Cambridge, when I started thinking, “I wonder, could we try to have this particular,” as you put it, “more targeted goal?”

My thought was that maybe we could reach more people if we developed a device that was much cheaper, so it could be affordable by people who were using it. That could avoid forcing a situation where you’ve got governments purchasing it, or governments who may mean to purchase it in one administration and then that falls through in the next one.

So, the two projects have some ideas of how we might collaborate, but I do think we’re doing completely different things. This is just very dedicated and meant to be in everyone’s hands, children and adults, and something they can purchase themselves.

Scott: Is this a Linux based device? What has it got for a software stack?

Cliff: If you’re going to develop something that you want to be able to sell for five dollars, you’re pretty limited in the type of microcontrollers you can purchase. We use one made by a company in Taiwan named Generalplus that builds them, typically, for audio devices and children’s learning toys and that kind of thing.

It has codecs built in that are optimized for voice compression, as well as a lot of the other basic support we need, like USB and easy access to flash memory through a MicroSD memory card, which is what we use for audio storage.

A lot of work went into finding the right microcontroller, and now we’re doing some pretty low-level assembly and C coding on it. They have a decent library, and they can compile C down into their own assembly language, so we develop largely in C. We don’t really have much of an operating system.

The underlying stuff written in C interprets what is basically an XML-like code, although we flatten it out a little bit more, so in the end, it’s not XML. It interprets metadata that drives the audio, so someone can mark points in the audio where there are page breaks, line breaks, or even word breaks, as well as when a word should be hyperlinked to a definition. It’s all based on metadata that the author can produce.

And the way an author might produce it is probably going to be through a GUI app on a Windows XP machine, because that’s what most of the nonprofit staff and the government workers in these district offices have at their disposal. They don’t have the Internet, in most cases, but they do have a basic Windows XP machine that they use for word processing and that sort of thing.

So, they would use a software application to mark the locations in the audio of these breaks, and the output of that application is some XML that is then fed into the device, and the metadata from that XML drives how the buttons on the device will jump around the audio.

The operating system itself, or the audio user interface, is written in the exact same language. It’s just as if it was another piece of content, only it’s the one that pops up when you turn the device on, and it provides access to controlling the volume and the home button, for instance.

By splitting up the operating system that way, we make it much more extensible and available for more developers to play with and to do new things with. For instance, we have C developers in Ghana, our pilot country, who want to get involved in the low level code. There are going to be many more developers out there who may want to experiment with creating new operating system features, and they’ll be writing XML to change how the audio UI works.

Scott: Is the source code that you guys have written available under an open license?

Cliff: Yes, we will be making all that stuff available. We don’t quite have it all posted yet, but we’re putting everything we do out under an open source license or an open content license.

Scott: Talk about the thinking behind that. There have been some projects that target very poor populations to provide pedal-driven or solar-powered water pumps, for example, and some are for-profit.

What is your thinking behind the advantages of having the source code available under an open source license? Are you hearing ideas from people seeing if they can do X or Y or Z?

Cliff: We are not a very well funded nonprofit, because we are very new and we started from scratch. We didn’t bring on big corporations to make contributions and sit on the board or that kind of thing, which is a very different sort of model. And so, in this stage where we’re very small, the best thing we can do is to make it easy for anyone to be able to help.

We are also very, very particular about who we hire. We do pay contractors for some work, like an electrical engineer and a mechanical engineer, and while we do get good discounts from them, we really do lots of interviews before we pick the right one. But even so, we know that the people working on this stuff may not have the absolute best idea, and they’re willing to agree to that possibility as well.

So, by publishing everything we’re doing, we are allowing people to say, “Wait a minute. I think you might be able to do something cheaper if, instead of using those several components, you used this one component.” Or the same thing with software is, “I think, we can probably make that code more stable if we write it this way.” And so, whether it’s around hardware and mechanical stuff or software, by opening things up, the best ideas are going to be much more likely to get integrated into the project.

Scott: I’m sure that the gulf between the people working on the project and your projected user community can cause a pretty significant disconnect at times. Can you talk about the development process and how you avoid making bad assumptions about how your target audience will interact with the device?

Cliff: Sure. One of the things we are doing, assuming we can raise enough money to produce the devices in time, is that we have a pilot program planned in October. We hope to take 100 of these units out to a very small, very remote village in Ghana.

This is one of the villages where we have already taken out an earlier prototype and gotten some feedback on it, but, this will be the first time that we’ll have completely functioning devices out in people’s hands. We’ll be there for several weeks, and we’ll have people there to help them with issues that come up, but mostly to observe what they like and what they don’t like, what they have trouble with, and that kind of thing. That’s where we’ll probably discover a lot of places where some of our assumptions were wrong.

So far, we’ve found that people are able to figure out a lot of this technology. Part of that is because where these people live, there are a lot of radios, so they are already familiar with some portion of the technology. They are accustomed to taking a device that requires batteries, turning a switch, and having sound come out of it, turning a knob to change it to a different station, and another knob to control volume. That has been around for generations, and everyone understands it.

Cassette players are another important analogy here. Not everyone has a cassette player, but in a lot of these villages, there is at least one, so it is not really a foreign concept. The idea of being able to start and stop recorded audio, and to fast forward and rewind is already there.

Because of all that, we don’t have quite as far to go as you might imagine in bringing people up the technological learning curve to this device. That said, there are definitely some changes that are going to be similar to people in the US or Europe going from VCRs to TiVo. That’s the kind of jump that people are going to need to make in order to use this device.

That won’t be true everywhere in the world. In places where people haven’t seen radios before, we’d have more of a challenge.

Scott: How did you get started with this?

Cliff: I worked in the software industry for a long time, including a stint at Microsoft for a few years, and then I transitioned into consulting, primarily around helping businesses decide when it was to their strategic benefit to open source their software.

During those consulting years, I had the flexibility to spend a lot of time reading up on other things. I spent a lot of time reading about global poverty and disease issues, and I also volunteered for a few advocacy organizations.

We’d sometimes go to Washington, DC and lobby members of Congress, and that required me to study all of the different foreign aid, micro-credit, education, and health programs, so I could advocate for one over another. In doing that, I felt like I got a good perception of what seemed to work and what didn’t.

One of the things that I began to feel really strongly about is the importance of education and long-term solutions to poverty. That led me to start thinking about these kinds of solutions, and as I mentioned before, I was interested in how OLPCs could be more useful.

I was planning a trip to a very rural part of Ghana for six weeks, to get beyond the political advocacy side and the big money politics in Washington, to see what it looks like on the ground and what works and what doesn’t. In planning that trip and working with OLPC, those two things kind of folded.

Looking at the Ministry of Education’s budget for Ghana, I realized there was just absolutely no way they were going to be able to afford One Laptop per Child in the next five years. And so it was sort of that kind of feeling that made me think, “Let’s see what we can do to get a more targeted solution out to more people sooner.”

Scott: It’s interesting, because there are all kinds of different programs. Some of them are buying mosquito nets. Some of them are trying to get companies to develop vaccines that wouldn’t otherwise be profitable.

I don’t want to put words in your mouth, but you seem to have zeroed in on the idea that if there was better education, and if people had more knowledge about things like how to keep their children from dying of dehydration or how to improve their crop yields, then maybe some of those issues would solve themselves.

While all the other things are important, that by itself would probably make a very big difference. Have I got that right?

Cliff: Yeah, I think, so. I remember talking to a senator, when I was advocating for a particular program. I was with a couple other people, and one person described the importance of supporting the Global Fund for HIV/AIDS, TB, and Malaria. And then someone else described the importance of supporting micro-credit programs. And then I described education issues and the Education for All initiative.

And while I was describing it, I said, “You might think of this as sort of short term, medium term, and long term. You have to deal with the health issues. You have to provide people, once you get past the health issues, a way to improve their own near-term economic situation, by starting up their own business and that kind of thing, with micro-credit. But, if you don’t deal with education, you’re always going to be dealing with those other problems. The health problems will never go away without improving education.”

I have always felt that I would rather focus on a long term solution. Even if I didn’t see the ultimate results immediately, that was more important to me. But, yet, we also need organizations like the Bill and Melinda Gates Foundation or some of the others focusing on the health issues; you just can’t ignore that there are people who need immediate assistance.

I think it was probably that kind of feeling that led to this device, which really tries to do both. Its initial concept was dealing with the long term, helping improve literacy skills in children and adults, because that is the basis of education. If you don’t have literacy, everything else is out of reach for you.

I should also say that it wasn’t my idea. It was when I tested this idea out in Ghana for six weeks that I learned from the local nonprofit organizations and the local government agencies there that this same device could be really useful in helping them reach more people with immediate information. It can inform them about things like health, agriculture, or how to start a small business.

That’s how I really started to get excited about that idea, because it did address this all-encompassing vision of long-term education, but it also delivers immediate knowledge now that’s relevant and can save lives.

Scott: I’m sure that you must run into skeptics, as I know that OLPC has, with people saying, “Look, electrical gizmos aren’t what the poorest populations in the world need. These are going to get turned into doorstops.”

What makes you really feel like you’re onto something that’s really got legs and is going to make a difference? What do you say to the inevitable skeptics?

Cliff: The first thing I would say is that we know we don’t have the absolute perfect design right now. We are taking small steps, one at a time.

There was a point, really early on, where I thought we would design a prototype, take it to some developing countries, get some feedback on it, and then we’d gear up production for 100,000 units. I figured that we’d maybe do a small pilot with the first 1,000 units, and then kind of keep moving from there. Within a couple of months, I realized, “Wait a minute. We’re not going to get this right in the beginning. There are going to be major aspects of the device that we are going to need to tune until we get it right.”

We’re working with people in Ghana and India on a weekly basis, learning from their feedback on feature trade offs and that kind of thing. Still, we know that we aren’t going to get this right the first time, so that’s why we’re going to do a series of pilot programs.

That’s why I’ve told our electrical engineer and mechanical engineer that when we have a tough choice, we just aren’t going to worry about it too much. We’re going to pick the choice that we think is best, and we’re expecting that we’re going to do a redesign after our first couple pilots. And so, in other words, we know that the way to do this right is to put it out, get real users using it for weeks and months, and then move on from there.

As far as the big picture, we’ve learned a lot from OLPC. They’ve been criticized so much, and much of it is unwarranted, even if some of it is warranted. But they were out there pioneering the way, and it would be crazy for us not to learn from what they’ve done.

I learned a lot when I saw these government promises fall through, that Nicholas Negroponte went out there and got all these governments to say they were going to make all these orders and then things fell through. There were also the competitive issues with AMD and Intel and OLPC. There’s a lot to learn from there.

I’ve done my best to pay attention to all of that, and when we’ve had for-profit companies show up and say, “We’re kind of interested in what you’re doing, “we’ve embraced them and said, “Well, let’s see how we can work together.”

Scott: I see that we’re coming near the end of our time. Is there anything that we haven’t touched on that you wish we had?

Cliff: [laughs] Well, the biggest thing on my mind these days is all about fundraising. When you’re a small nonprofit startup, you start with funding the entire thing on donations from family and friends. What’s been interesting is that now it’s starting to spread, and only in the last few weeks, I’ve started to see donations come in from people whose names I don’t recognize. [laughs] And it’s a really nice feeling.

But, we are not funded by any foundation, and the reason for that is that starting an organization to do some good in the world is a difficult thing to do, because very few foundations will even look at you, unless you are already a 501(c)(3) tax-exempt organization.

Getting that status takes roughly six months, and it was only about two months ago that we got that letter. And now a lot of international organizations like the UN, the World Health Organization, and the World Bank have contacted us and said, “We’re really interested in what you’re doing. Can you tell us more about it?”

The moment that I say that we haven’t got it in the field yet–that we’ve got a prototype and a pilot test planned–they all say, “Oh, OK. Good to know. Call us back as soon as you’re done with your pilot.” [laughs]

It’s interesting that a lot of these large organizations, and even some foundations, are a bit risk-averse. If you’re not handing out food or bed nets or any of that, it’s a tougher thing for people to get their minds around. The donors that we do have donate to us is because they feel like their donation is being leveraged enormously, and all of this R&D is going to end up in millions and millions of these devices that will be sold at cost to people.

And even though there’s a bias towards wanting to give something away, that’s not really a sustainable way to deal with the problem of about a billion adults who can’t read.

Scott: Thanks a lot for taking the time to chat.

Cliff: Thank you very much. It was fun talking to you.

Bookmark this: