• E-mail this page
• |  Print this page
• | 
• | 
• | 

Hold The Steak, IBM--Let's See Some Sizzle

Sure, the intellectual heft was there. IBM's VP of emerging technology, Rod Smith, as colorful a character as you'll find at Big Blue--he shoots from the hip and showed up in two-tone cowboy boots--flew in from North Carolina. And John Patrick, a consultant who worked for IBM for 34 years, made the trek from Connecticut to the Left Coast. He developed the ThinkPad and retired as IBM's VP of Internet technology in 2001. IBM started AlphaWorks in 1996 to get underfunded and officially unloved technologies from its Watson lab in upstate New York to market by posting them to the Internet, Patrick recalled. It modeled the program after Lockheed Martin's Skunk Works, a coterie of aircraft designers working in secrecy and cosseted from corporate bureaucracy.

"Back then, the Internet was a threat," said Patrick. At the time, IBM was generating nearly $1 billion in annual revenue from proprietary networking technology, clearing most of that as profit. When the company toted up revenues from Internet technology--TCP/IP stacks, routers, and the like--the total came to $54 million--for the whole industry. Almost none of it was profitable. Yet somewhere, the light went on that this was the future. Patrick worked in the '90s with Irving Wladawsky-Berger, an IBM VP and its resident emerging technology sage, which probably helped. Today there are about 200 technologies on AlphaWorks, and IBM has posted nearly 700 projects there during its lifetime. More than 175 AlphaWorks technologies have found their way into products. Among the most popular areas of the site: grid computing, Java, RFID, the new Cell processor, and autonomic computing.

This week, IBM researchers posted three new projects. "Deep Thunder" is a weather prediction app that can zero in on patterns in areas as small as a square kilometer. "Web relational blocks" lets users create online apps by dragging and dropping commands in a browser without writing code. "Adieu" could let average Joes quickly create Web services without programming know-how. And VP Smith is working on an app called "QED Wiki" that can create mash-ups of business apps with a few lines of HTML code.

The problem was, no one particularly topical showed up to peek inside IBM's innovation machine. On hand for the conference-room champagne pour were a motley crew of second-tier tech reporters and analysts--the kind of people who look like they've spent too much time behind a PC packing in the junk food and letting the grooming slide. Lots of bowl haircuts and worn-out shoes. A middle-aged guy in a T-shirt scowling in the front row looked like the bearded version of Saddam Hussein in an outtake from those trial photos. And the low-res tribute video IBM's managers screened to kick off the afternoon didn't help raise the buzz factor--a few more production dollars wouldn't have hurt.

That, and IBM's announcement underwhelmed. On the 10th anniversary of building a public bridge to its research labs, IBM unveiled a new site called "AlphaWorks Services." Instead of downloading the software prototypes, interested users can test them directly over the Net, letting IBM stream out updates and collect feedback more easily. Seemingly most important, the move lets IBM grab some of the Web 2.0 pixie dust--that moniker, as well as "software as a service," arose a couple of times during the proceedings.

IBM has shown time and again it's capable of thinking and actions that lead the computer industry in new directions. Just this week, the company said it would start posting its patent filings to the Web for peer review in a break with widespread practice. IBM hopes the move, which would let competitors see what it's working on years in advance of patent awards, will quash some of the abuses of the U.S. patent system by tying authors more transparently to their corporate backers and asserting a policy that bars patents for fuzzy business methods without technical merit. It has also convinced companies including Microsoft, General Electric, HP, and Intel to submit some patents for online, open source-style peer review next year. The move poses some risk to IBM, the No. 1 patent recipient in the United States last year with 2,941, but could pay dividends by raising market pressures against dubious claims. IBM assembled a group of patent experts to hash out details this spring on a wiki that's available online.

"Nobody ever got fired for buying IBM," goes the old saw. But which IBM are you buying from? The one whose personnel are steeped in some of the deepest business, technical, and economic expertise on the planet, and whose systems run the world's engines of government, science, and commerce? Or the IBM that seems to continually reach for a patina of freshness, only to fall into a familiar drabness that makes its mechanisms impregnable to the computer-using public? IBM officially gave up on the consumer market last year when it sold its PC business to China's Lenovo (most of that was corporate business anyway). Yet at a time when Google sets the standard for cool places to work, IBM's older employees are planning a Supreme Court appeal of their suit against the company for scaling back its pension, and the general rate of U.S. innovation is under scrutiny, a window into IBM's works could generate more goodwill. A little more marketing sizzle could be a first step in that direction.

« Consumers Hanging Up On Indian Tech Support | Main | Consumer Tech In IT? Why Not? »