General — Page 6 of 246 — LIRNEasia


I have been asked several times about the feasibility of national online freelancing platforms. There are a few in Sri Lanka. According to this report, it appears that a platform has emerged in Myanmar too. First prize goes to Honey Mya Win. She founded her startup with her programmer sister less than a year ago after quitting her job with a Chinese telecom firm.
According to information obtained from the Facebook advertising portal, we calculate that 29 percent of Myanmar’s population has a Facebook account. I wanted to cross check from our 2016 survey. We found 35 percent of the mobile users were also Facebook users. The base here is smaller (not the entire population by telephone users in the 15-65 age group). So obviously, our number has to be higher than 29.
I was working on some comparative numbers. Most of these are recent and from reliable, credible sources. Interesting insights. Most people think Facebook use is a subset of Internet use. But in SE Asia, Internet use is always lower than Facebook use.
Micro and small enterprises play an important role in developing countries. Any improvement in their performance is likely to have a significant impact on the poor. Many studies have been conducted on the impact of ICTs on MSME performance. Research Fellow P. Vigneswara Ilavarasan, who has been conducting research on MSMEs for many years, led a team which conducted a systematic review of this research.

Rule of thumb is also an algorithm

Posted on June 28, 2017  /  0 Comments

As part of our big data work, we’ve been thinking about the opacity of the algorithms we use. The pretty picture and tables that result from the research are persuasive, but if people wanted to know how they were derived, it would not be easy to explain. But then, we have to always think about the alternative. A method may be familiar and may have been used for decades if not centuries. But that does not necessarily make it fair.
Been thinking about AI for a while but realized there was nothing on the record. It’s good to have some record of what we are thinking about, as illustrated by the recent tweet I sent showing our first post following the launch of the iPhone. What is artificial intelligence today? Roughly speaking, it’s technology that takes in huge amounts of information from a specific domain (say, loan repayment histories) and uses it to make a decision in a specific case (whether to give an individual a loan) in the service of a specified goal (maximizing profits for the lender). Think of a spreadsheet on steroids, trained on big data.
When I first came across O3b in the Pacific, I asked a lot of questions about latency. Because the answers were right, I’ve been recommending O3b type solutions to people who want satellites as part of the solution to broadband connectivity problems. O3b went from four to twelve medium-earth-orbit satellites, serving niche markets that could not be served by fiber. Its weakness, if any, was that it could not serve the northern latitudes. Now Greg Wyler, a founder of O3b, is seeking to fill that gap with a massive constellation of over 700 satellites in a new system that has the financial backing of Intelsat and was just licensed by the FCC, OneWeb.
Hearing the many reports on prosecutions under section 66(d) of the Law that was enacted in 2013, I went back to my files. In the extensive comments we provided there is nothing that refers to the offenses sections. The offenses chapter is peculiarly drafted. Section 65 is similar to what is found in any law that requires a license to be obtained for a specified activity. Section 67 is again a necessary section, specifying the penalty for using equipment without a license.
The first time this happened was when Pacific Century Cyber Works (PCCW), controlled by Richard Li, acquired Hong Kong Telecom. But that 2000 adventure did not end well. PCCW’s stock price tanked. Seventeen years later, things are different. The real story may not be a state-sanctioned infusion of private capital into a 100% state owned infrastructure company, but the first move by firms in the upper layers to take over entities in the infrastructure level.
Dr Saman Kelegama, the head of the government supported economics think tank Institute of Policy Studies, has suddenly passed away while on official business in Thailand. He was at the helm of IPS for 22 years under different administrations which is quite an achievement and one that many will talk about. I worked with him on the India Sri Lanka Comprehensive Economic Partnership Agreement joint study group back in 2003. We were both attacked by the petty opponents of any trade agreement with India. He was appointed to head trade negotiations by the new government.
Apparently, the mechanism to shut off the screen when an iPhone was brought close to one’s head did not recognize black hair at the outset. The book review has such nuggets. The book must have much more. In fact, although it would eventually emerge as the gleaming quintessence of the collaboration between the Apple co-founder Steve Jobs and Apple’s design magus, Jony Ive, Purple could seem like a nightmare of overwork, insoluble technical tarballs and political infighting. “You created a pressure cooker of a bunch of really smart people with an impossible deadline, an impossible mission, and then you hear that the future of the entire company is resting on it,” Andy Grignon, one of the iPhone’s key engineers, has said.
LIRNEasia started working on broadband quality of service experience as long as as 2007. The early part of the story is narrated in this five minute video. One of the earliest conclusions we reached was that speed was just one dimension of performance. We communicated these findings to TRAI as long ago as in 2009. It’s a pity that speed is the only aspect mentioned in the title of the consultation paper for which comments are due by the 29th of June.
Looking for something in my files, I found this conference paper that is almost 10 years old. The organizers pressured me to write it but then they did not keep their side of the bargain and publish the proceedings. It requires a few hours of work to make it up to date. The basic structure is fine, and could even be used to assess the WTO compliance of other countries that have made telecom commitments. Pity it never saw the light of day.
A friend of mine who served as comptroller of a large diversified conglomerate once told me that there was a clear difference between the business units within the conglomerate. Those which produced items for export were fixated on efficiency and were nimble in terms of responding to market demand. Those who produced goods and services for the local market were fat and lazy. I was reminded of this when I read this report about India trying to do something about government capacity. Unlike in government entities that serve citizens, those that engage with the outside world cannot be fat and lazy.
In a parable I worked up in 2012, I speculated on the possibilities of joint ventures between Internet companies such as Facebook and the last-mile access companies to enhance the user experience. Some details of a dispute in South Korea shed light on the problem: According to SKB, there were initially two ways to connect to Facebook in Korea: via a direct connection to Facebook’s server in Hong Kong and via rerouting to a local cache server in Korea operated by local telecom provider KT. The cache server is used to save online content locally in temporary storage, called a cache, and in turn improve the connection speed for accessing foreign internet services. Facebook currently pays KT to use its cache server. SKB argued that Facebook deliberately cut off its link to KT’s faster cache server last December and has since been clashing over network maintenance issues.
When I was living in US Midwest, Delphi was a familiar name. It was a company that employed thousands to make original equipment for vehicles. But now, according to NYT, Delphi is positioning itself to be a player in the data business: Delphi hopes to create a new business that can gather vast amounts of data from vehicles — about how and where they go, how they’re driven and how they’re running. The company then envisions selling insights drawn from the data trove to automakers, insurance companies and possibly even advertisers. A driver who frequently drives to Starbucks locations, for example, could be targeted with Starbucks coupons via email or text.