News by Computers



By TAN Kee Wee
(MediaCorp 938LIVE’s Money Talks, Friday, 9 November 2012, 7.55 am, and Saturday, 10 November 8.35 am)

The internet is a great place to search for almost anything. However, the amount of information available can be overwhelming.

Not to be outdone by this, a 17-year-old British student, by the name of Nick D’Aloisio, has found a solution, at least for those who want to keep up with the news.

He has created a smartphone app which provides summaries of news stories found online. The free app, by the name of “Summly” was launched last week and became, within hours, one of the Top Ten downloaded apps

“Summly” will condense news articles to about 400 characters, or less than 100 words. Future versions of the app could allow users to read the full article just by tapping on the app.

This ability to read the full article could lead to a pay-per-view business model. In future, readers could pay to read one newspaper article rather than pay for the monthly subscription. It will be very much like how we buy a single song for 99 cents now, rather than paying more than $15 for the whole CD album.

Its money-making potential is the reason why Summly has attracted investors like Rupert Murdoch and Hongkong billionaire Li Ka-shing.

The instant popularity of Summly could be due to it being the first-of-its-kind. Currently, there are services that provide news summaries. But these summaries are done by humans, and they are expensive. Summly uses computer algorithms, or software, and they are cheaper and faster.

Replacing humans with computer algorithms is not new to the investment industry. It has been estimated that between 40 to 80 percent of trades in the US$50 trillion global stock markets are already being carried out by computer algorithms.

Last week, UBS, Switzerland’s biggest bank, became the latest giant to replace human power with computing power in another area of their investment division.

Instead of paying their credit default swap traders million dollar salaries, UBS has invested hundreds of thousands of dollars to install computer algorithms that trade using mathematical models.

Besides reducing costs and other benefits, the move would phase out the current practice of making deals over the phone. It is very much like how stocks used to be traded before computers took over.

Of course, there are dangers of introducing computerized trading into the US$30 trillion credit default swap market. Critics have said that it could lead to violent price swings. They say that the flash-crash of May 2010, when a computer error brought US stocks down by 9%, is an example of what could happen.

Some have claimed that the financial crisis of 2008 was the result of computer models misleading investors into thinking that all was well. These fears are not far-fetched, especially when different computers usually come up with the same logical solution.

Despite the weaknesses, the attractiveness of replacing humans with computers cannot be ignored. It’s the same with the Summly app. The cost savings can be huge if its principle software is adopted by the news and research industry. Many jobs would be lost.

But most importantly, if we investors are fed the same computer-generated news summaries and investment advice, our views would become one.

A potential disaster exists when we combine computerized trading with computerized news summaries. There will be no diversity of investment views and investment strategies.

We would be like the group of farmers who chose to plant one strain of food crop because it is the most convenient, or we are too lazy to try another strain of crop. With no diversity of crops in the farmlands, when a disease strikes, all crops will be wiped out.