PeterMe asks what Big Industrial Research is for: after all, he points out, lots of innovation comes from little companies and pairs of clever people putting stuff together--while MSR brought us Clippy.
peterme.com: Research and Development in Interaction Design
Now I have the good fortune to be a friend of arc Smith, to have been involved in parts of Remail, and to have an advisor who used to work at a prominent industrial research firm or two.
I've certainly heard a number of cynical takes on the role of research labs:
There are more optimistic flavors: that having a research lab encourages innovation in a company that is large enough to start slowing down. Or that having a research lab means that you have interesting internal presentations on the state of the art that can help drive other development.
But let's assume it's all about the product.
Peter skips over several important transitions: how does the research get motivated? how does the research get transitioned into a product?
The latter question may be what actually kills research more often than other times: Bell Labs, famous for getting smart people together to do clever things, never solved that problem; neither did Xerox PARC (inventors of the mouse, the GUI, ethernet, and more cool stuff than you can imagine).
Times are changing a little. Research is now usually forced to orient their products at the company. For example, I know that some Microsoft researchers have explained that their own time is paid for by the company out of research funds, but any developers and hardware they get need to be provided by product groups that are willing to invest.
At IBM, to the best of my knowledge, research teams get some seed money from the company, but then need to contact product groups for matching funds: $1 from a product group will be matched by another dollar from the company.
This leaves groups in a variety of positions. One team at IBM had held onto the Little Red Dot long after it stopped being research because they needed the income stream from the laptop group.
The Remail team at IBM pushed hard against the Notes group to get money and interest in Remail. They succeeded, and built a working prototype, which they passed on to dev. I believe it's scheduled to be a product in a year or two.
Other groups keep a mix of 2-year horizon and 5-year horizon researchers around, trying to balance tomorrow's software with farther-future stuff.
This seems like where research really can shine: sure, it has a lot of failures. But it can also think in ways that small-business innovators can't. Peter listed a bunch of clever tools put together by small teams and innovators--but they don't have the access to large organizations to build big things.
And I think there's room for both. We need CmdrTaco to write up Slashcode (small), and we need someone to provide a few dozen terabytes for Marc to archive and cross-reference Usenet (medium) and we need someone with chip fab plants and machine rooms to build little red dots and advanced graphic cards.
February 16, 2004 11:22 AM | TrackBack | in Design