IBM's Web-based intermediaries
Jim Spohrer. 1.*
- unedited transcript -
Okay, I'm Jim Spohrer from IBM Almaden research. Doug has asked me to tell you a little bit about the WBI [pronounced 'webbie'] technology. I'd like to start with a question: can anyone guess why we decided to call WBI WBI? Yes, in the back? Yes, that's correct. [laughter] It stands for IBM upside down, very good!
So um, the WBI technology, how many here have written a java applet
or program? Ok, about a quarter or third or so. If you're unfamiliar with
java, then I'm going to give a very brief technical definition and then
I'm going to get into some more application-oriented stuff. So think of
WBI as a web programmable proxy server that can exist on the client proxy
position or server that can monitor the flow of information and edit and
customize it. Why WBI? Information is flowing around us in the world; it
is flowing everywhere. And whenever information flows, there is opportunity
to add value to that information and so for example, if a person is reading
a web page and doesn't understand a word, there is an opportunity to come
in and provide a definition for that word.
Or if you're reading a web page and you see the word 'Intel' then you can go see what the latest stock price is. This is just some of the stuff you could do if you're monitoring the flow of info and thinking about how to add value. And I predict that by about 2010, paper books, as we know it will go away; we won't have any paper books any more. Why won't we have paper books? Paper books are inefficient because they don't have intermediary capabilities built in. If you're reading a book and you want to have the book read to you; if you're reading a book and you don't know the meaning of a word; if you're reading a book and you want more information, intermediaries are the way to go. Now I'll tell you a little bit about intermediaries as we go along but I remember the early days of designing educational software, which was something I did in the distant past. When we were building educational software, one of the hardest buttons to build - you're trying to design educational software and then you try to design this little button for more information. And in the old days - even in CD-ROMs - when you push that button, you have to design all this extra stuff. But in the connected world, the 'for more information' button became, from the most difficult button to define, to the easiest button to define. And that goes to show that the paradigm shift that can happen. Things that were very hard in the past suddenly became very, very easy if you had the right technology.
Scattered through this talk will be a number of URLs. If you want WBI it's free, you can download it from Appleworks or from various websites, but you can also download it from a number of various applications for WBI as well. oh, here we go.
You can think of WBI as a kind of smart pipe that information is flowing through. And that smart pipe can monitor the information going through it, it can edit it, it can generate it, it can modify that flow. And remember, the flow of information on the web is going both ways; it's not just going down from the server, it's also going up as users fill out forms and so forth.
Well, what can you do with a smart pipe? Well, one of the biggest applications
that IBM envisions is transcoding. As we have more and more pervasive devices
that have smaller screens, black and white screens, you want to take information
and run it through transcoders. Transcoders can take color images and turn
them into grayscale images. If you've heard about, for example, Yahoo Anywhere,
you want Yahoo on your cell phone. Transcoding is the technical term we
to that translation of the information for a different pervasive form factor
device. There's a lot of other things you can do though, you can translate
the content into a different language. One of the things you can do is
make it more accessible, you can run it through text-to-speech systems.
There's just lots of information that's flowing around. And one of the things that I do as a hobby now is I listen to these radio stations that are constantly advertising dot com companies. I don't know if that's just the rest of the world as well; it probably is - but here in the silicon valley you get all sorts of advertisements: dot-com this and dot-com this. If you're not converting from bricks to clicks, you're going to be extinct. So anyway, I hear about a lot of companies, and more and more of the companies that I keep hearing about are what I call intermediary-based companies. What do I mean by that? They're adding value to somebody else's stuff. And it's very interesting if you look at Novell has 'digitalme,' there's also ECode, there's even an article in the paper recently about what are called infomediaries; they're just simple things that'll fill out the forms for you. So if you go to Amazon, and you don't want to have to fill out the form every time, you go to another website and you don't want to have to fill that out, and enter your credit card information number over and over again, that's what an infomediary will do for you. Another example of a kind of infomediary application is print.org. These are interesting intermediary applications because they change the control point a little bit. Someone puts up a website, they're in charge of the content of the website. But what if someone has a comment? I agree/disagree? These are the sorts of capabilities that print.org or Thirdvoice can do. Intermediaries are that kind of thing. Because if you think about having a programmable proxy server, what the proxy server can do is get the request to go to the webpage, goes out and gets the webpage, but it can also go to another source of information, combine it with that information, and show you what somebody else wanted to say about that information. Guru.net is a great one. Every single word is hyperlinked. We termed this 'advisors' and if you go to our WBI webpage, there's a medical advisor, so if you're looking at a medical page and there's a medical term, we take you to one of the best online medical dictionaries on the web to see what the definition of that word is. There are also a lot of companies springing up online like, AreYouSure - And this is the comparison-shopping. And remember, this company didn't have to build ebay or amazon.com. What it had to do is say, because these exist out here and because I can imagine someone wanting to do this thing, and I, the intermediary, can do it. So if you really want to think about it, it's sort of like the middleman. I remember in the early stages of the evolution of the web, there was a lot of talk about the middleman goes away. No! The stupid middleman goes away. The ones who don't add any value go away. There's always a way to add value to somebody else's stuff! So it becomes one of these - what I call - 'enabling technologies.' If you have these technologies, there's just so much more value that you can bring into other people's work. I could go on and on, but in the interest of time I'll go on.
This is for some of the developers and programmers in the group. This is the architecture [of WBI]. The interesting thing that we're thinking about is http for the web is just one protocol for web information flow. There's all sorts of protocols, you have your WAP protocols, ftp protocols, there's even protocols for your television sets, you know, the signals and information for your television. This intermediary architecture notion is very general. How many of you know about tivo? Just about everybody of course. Tivo is an intermediary play. Maybe it's just that I've got a hammer and everything else looks like a nail, but it's really an intermediary play and what the intermediary value is storage. So what you do is you're watching a TV show but Tivo stores up the programs; you can pause real-time television - that's the sort of stuff you can do. This is, in Doug's terminology, one of those important infrastructure technologies. This is one of those capability infrastructure pieces that are coming into the world now. We're making this technology freely available to the world as we can at Alphaworks, we've got a lot of requests to make this more open and that's one of the reasons why we're here; we'd like to hear from other people who are interested in working with us to create a more open story. It looks like that was my last slide, so back to Doug. But just one more thing before I go away, if anyone disagrees with the fact that books go away by 2010, the thing that I said - I can't remember if I said it to this group - was when's the last time you touched an LP record?
AUDIENCE: I do.
SPOHRER: Aha! So there will still be people that still touch books in 2010 but most of us won't. Most of us here haven't done so in 5 or 10 years. We've got them sitting out therein boxes in our garage or we've gotten rid of them. Yes. Question.
AUDIENCE: A lot of the software that is out there is freely available but is not open source, isn't it?
No, and I don't want to mislead anybody about this. We've got a lot of requests about if somebody says, I want an open source intermediary, I can give you pointers where there are open source intermediary codes. The technology that we have now is not open source. We've got a lot of requests from a lot of groups that would like to work with us on this possibility. And just so you know IBM's policy on the matter, we don't make things open source until a lot of customers have adopted the technology and they've said it would be beneficial for it to be open source. So we want serious people who are willing to adopt the technology as it is today, and start working with it so it will show some credibility to their request, not just somebody saying, "gee it would be good to make it open source" and then we make it open source and they walk the other way and don't do anything with it. So that's what we're trying to do: find credible partners that will work with us to really ensure that if it's made open source it evolves and there's the right structures there and I can't guarantee that; I can only speak about what the technology is and the process that we go to reach that. Yes. Question?
AUDIENCE: if books go away by 2010, do you think that the replacements will be covered by shrink-wrapped licenses and if so, do you think that therefore all copyright protections like fair use and first year will also go away by 2010?
That's a very great question. This is really a co evolution question, to relate it back to Doug's discussion. This is where technology can evolve to take us some places but we really need laws to get us there. I'm not going to predict how the laws are going to evolve because it's hard to do that because business models change. And it could be business models that change; we're seeing this with the mp3 stuff right now and it's really hard to predict. So what I'm trying to say is that there's going to be a way for books to go away and it's going to make a lot of people billions and billions of dollars. And how that involves in terms of laws and technology, I could speculate but don't want to speculate right here.
AUDIENCE: The direction of the laws is to do away with fair use and first sale; that's the direction we're moving.
AUDIENCE: Jim there is a company that's doing a lot of work in this area right now that has been very successful. It's called iCopyright, and I don't know if you've seen it. ICopyright.com
SPOHRER: Yes. Al?
AUDIENCE (muffled): Yes, I was just wondering. Does IBM community allow you to download from an Appleworks site? [can't really make out]
SPOHRER: Did I say that? Alphaworks, I meant. Sorry. Very good Adam. My son follows me around and corrects my mistakes. [laughter] Ok, with that I'll turn it over to Doug. Thank you all. [applause]
Above space serves to put hyperlinked targets at the top of the window