Robot Reporters

Quarterly business earnings reports are dull and boring to read. That makes them ripe for automation. The Associated Press has been using an automated system since last year to write its stories and few readers noticed. AP implemented the system six months ago. It now publishes 3,000 such stories every quarter and that number is poised to grow.

AP says the automated system is now logging in fewer errors than the human-produced equivalents from years past. Of the estimated 3,000 such reports each quarter, about one hundred will have an added human touch, either by updating the original story or doing a separate follow-up piece.

The giveaway is that there is no byline and at the end of the article we see, "This story was generated by Automated Insights." The Wordsmith platform also generates millions of articles per week for other companies, such as Allstate, Comcast, and Yahoo, whose fantasy football reports are automated.

What's in a Name

Twitter - A small group of employees from Odeo, the San Francisco podcasting startup where Twitter initially began, had a brainstorming session. They were trying to come up with names that fit with the theme of a mobile phone buzzing in your pocket with an update. After narrowing down the options (which included Jitter and Twitter), they wrote them down, put them in a hat, and let fate decide. Fate decided on Twitter as the name was literally picked out of a hat.

Yahoo - Founders David Filo and Jerry Yang started what would become Yahoo when they were Ph.D. candidates at Stanford University. The project originally consisted of categorized lists of favorite links on the web, which made its original name, “Jerry and David’s Guide to the World Wide Web,” at least accurate if not so catchy. Yahoo is actually an acronym for “Yet Another Hierarchical Officious Oracle.” According to the company, the team chose the name for its definition: “rude, unsophisticated, uncouth.”

Gigabytes, Terabytes, and Petabytes

One gigabyte of storage cost $228 in 1998 and costs about 88 cents today. A gigabyte holds about 7 minutes of a HD TV movie.

A terabyte is 1,024 gigabytes. The first terabyte hard drive came out in 2007. In May 2009, Yahoo! Groups had 40 terabytes of data to index.

There is an even larger scale these days, petabytes. One petabyte is 1,024 terabytes. A petabyte holds 13.3 years of HD TV movies. To provide more perspective, Google processes an average of about 20 petabytes of data per day. Finally, 50 petabytes is equal to the entire written history of mankind from the beginning of recorded history. What's next? An exabyte is 1024 petabytes.

Health Insurance Myths

On the Fortune 500 list of top industries, health insurance companies ranked 35th in profitability in 2008; their overall profit margin was 2.2 percent. They lagged far behind such industries as pharmaceuticals,  profit margin 19.3 percent, railroads (12.6 percent), and mining (11.5 percent). Among health insurers, the best performer last year was HealthSpring, which showed a profit of 5.4 percent. “That’s a less profitable margin,’’ AP noted, “than was achieved by the makers of Tupperware, Clorox bleach, and Molson and Coors beers.’’

For the most recent quarter of 2009, health-insurance plans earned profits of only 3.3 percent, ranking them 86th on the expanded Yahoo! Finance list of US industries. Makers of software applications, by contrast, are pulling in profits of nearly 22 percent.