This image of pentacene, a molecule made up of five carbon rings, was made using an atomic-force microscope. Credit: Science/AAAS

Using an atomic-force microscope, scientists at IBM Research in Zurich have for the first time made an atomic-scale resolution image of a single molecule, the hydrocarbon pentacene.

Atomic-force microscopy works by scanning a surface with a tiny cantilever whose tip comes to a sharp nanoscale point. As it scans, the cantilever bounces up and down, and data from these movements is compiled to generate a picture of that surface. These microscopes can be used to "see" features much smaller than those visible under light microscopes, whose resolution is limited by the properties of light itself. Atomic-force microscopy literally has atom-scale resolution.

Still, until now, it hasn't been possible to use it to look with atomic resolution at single molecules. On such a scale, the electrical properties of the molecule under investigation normally interfere with the activity of the scanning tip. Researchers at IBM Research in Zurich overcame this problem by first using the microscope tip to pick up a single molecule of carbon monoxide. This drastically improved the resolution of the microscope, which the IBM scientists used to make an image of pentacene. They arrived at carbon monoxide as a contrast-enhancing addition after trying many chemicals.

The researchers hope that looking this closely at single molecules will give them a better understanding of chemical reactions and catalysis at an unprecedented level of detail.

The imaging work is described today in the journal Science.

To some engineers at IBM, the traditional approach to software analysis is far too inefficient. Data is collected and stored in a repository, and then software breaks off chunks of it to analyze. Some time later, the software spits out a result. But recent work at IBM is providing a better way: analyze the data as it's collected. The concept is called stream computing and it could revolutionize industries like finance, health care, and weather monitoring, where real-time data and analysis can help people make better, faster decisions.

In April, IBM showed off a system that could analyze the constantly fluctuating value of stocks. Now the company is working toward developing a product, called System S, that could be applied to any field in which numbers need to be crunched quickly.

According to a recent New York Times story:

I.B.M., based in Armonk, N.Y., spent close to six years working on the software and has just moved to start selling a product based on it called System S. The company expects it to encourage breakthroughs in fields like finance and city management by helping people better understand patterns in data.

...

Instead of creating separate large databases to track things like currency movements, stock trading patterns and housing data, the System S software can meld all of that information together. In addition, it could theoretically then layer on databases that tracked current events, like news headlines on the Internet or weather fluctuations, to try to gauge how such factors interplay with the financial data.

IBM has announced that it is developing a computer to compete on the game show Jeopardy! If the system can outwit human contestants--by comprehending and responding to a variety of questions more rapidly--artificial intelligence will have made a significant advance, the computer's developers say.

Back in 1997, IBM's chess-playing supercomputer Deep Blue famously beat then world champion Garry Kasparov in a highly publicized match. The event was symbolic in that it showed that computers could outsmart humans at a game once considered too intellectually challenging for a machine to master. Even so, chess is a game with well-defined rules and limits. Playing Jeopardy!, in contrast, requires a computer to deal with a variety of subject matters, from politics to pop culture, and to answer questions based on clues that involve analyzing subtle meanings, riddles, and puns. Jeopardy! might seem a lot simpler than chess, but for a machine, it's a far harder challenge.

According to the New York Times,

Under the rules of the match that the company has negotiated with the "Jeopardy!" producers, the computer will not have to emulate all human qualities. It will receive questions as electronic text. The human contestants will both see the text of each question and hear it spoken by the show's host, Alex Trebek.

The computer will respond with a synthesized voice to answer questions and to choose follow-up categories. I.B.M. researchers said they planned to move a Blue Gene supercomputer to Los Angeles for the contest. To approximate the dimensions of the challenge faced by the human contestants, the computer will not be connected to the Internet, but will make its answers based on text that it has "read," or processed and indexed, before the show . . .

I.B.M. will not reveal precisely how large the system's internal database would be. The actual amount of information could be a significant fraction of the Web now indexed by Google, but artificial intelligence researchers said that having access to more information would not be the most significant key to improving the system's performance.

IBM has already conducted some laboratory demonstrations of the program and still has some bugs to work out, such as getting the machine to understand the way that Jeopardy! clues are offered and what it should be searching for, leaving some experts skeptical that the computer program will vastly change the field.

According to the show's producers, an episode of Jeopardy! will be aired (at an unspecified date) pitting the IBM system--named Watson, after IBM founder Thomas J. Watson Sr.--against several human players.