Drew Cogbill | Thesis Blog

Technology for Developing Nations (1 December, 2006)

Eric Brewer, UC Berkley - TIER, TIER blog

Response: Most important ideas for me from Brewer: non-loss business, co-development, wifi for bootstraping.

Notes:
- Technology can make a difference
- Advocates for focusing on the rural poor and urban slums, as urban areas are moving forward due to market forces
- Development is: governance and technology
- Bringing wireless technology to an area can bootstrap on other infrastructure
- Wireless infrastructure is incredibly affordable, only dirt roads are cheaper
- non-loss business - sustainable but not profit generating business
- Technology solutions can be franchised from a third party loan to create a non-loss business
- River Blindness, Black Fly eliminator - used sensors to look for areas prime for black fly and sprayed - protects 30 million people from infection and makes 100,000 square miles now viable for farming
- Poor people spend lots of money, and even the poorest have a little disposable income, and they spend money on entertainment and communication, 7% of income in rural Bangladesh spent on telephony
- Mentions that technology could help in long distance money transfer ***
- Grameen Bank, Bangladesh- micro-loaning, began when noticed that poor friends paid back their small loans
- Grameen Telecom- microloans for village phones, a mobile phone woman in each village, a “super pay phone”
- Scaled with micro-finance- (***phone based money transfer, micro-financed, franchised)
- Make some project in the first six months, make shorter-term plans.  5 year plans are not ok with technology; things change too quickly.
- Develop something with a group and deploy- co-develop, co-deploy
- They have a $50 board to make a more efficient solar system
- Working mostly on long-distance wireless
- Mainly employed with rural telemedicine with rural eye clinics in India.
- These clinics serve areas without doctors.  They doctor video conferences in and tells patients things after conferencing with a ‘nurse’
- They get people to go get cataract surgery
- They’ve made these clinics scalable, they’re deploying 50 more

Literacy
- Voice is a big part of developing nation communication
- there has been no work on voice recognition for many languages
- Instead of goings at it typically, they’re going to do sound recognition - won’t work with translation but will perhaps work with user interface interaction
- Make the recognizer context sensitive, minimize the legitimate words that can come in
- They’re recognizing digits well with this method
- How do you get training data from illiterate people?  Using fingers to get people to say them, make people comfortable, use locals to interview, use a phone versus a microphone
Education
- cell phone game to teach english

Paper mentioned to look up: “The Case for Technology for Developing Regions” - E. Brewer et al. IEEE Computer, June 2005