Tim O'Reilly: Well, I find myself actually in a surprising amount of agreement with Jay, which I wouldn't have expected. But let's back up a few steps here, just to talk a little bit about the characterization of these issues. I became involved not as a theoretical issue around any of these questions, but really as a pragmatic issue. Many of you probably already know the history, but it's probably worth repeating.
I've actually owned a few patents at various points around software, so I wasn't knee-jerk about it. Somebody once said, "Oh, there should not be any property in this area." But I started to get pressure from my customers who were saying, "We're really mad at Amazon.com about the one-click patent. We'd like you to stop selling your books to Amazon."
I said, "Gee, I've got a lot of people complaining. Why are they so hot and bothered?" And so I wrote a letter to Jeff Bezos saying, "I think you're making a mistake. You're angering a core set of your customers. What's more my customers, the people who are building Internet technologies, are probably your key suppliers of technology. You know, you are not a technology company. You are a company that is benefiting from a community of people who are not playing by patent rules. They're playing by a different set of development rules (which Vint aptly summarized earlier on). It's a culture that has been extremely innovative, extremely productive for all of us as an industry."
This was purely between me and Jeff as a private conversation originally. I said, "I think you're making a mistake," and I used the colorful phrase "I think you're pissing in the well." And, I still believe that, for me, that is one of the fundamental issues.
There is a pragmatic choice that we are making as an industry right now and that choice is whether we want to continue letting the goose lay the golden eggs or whether we want to slaughter the goose and see if we can do a little better that way. And I personally believe it's just purely a pragmatic mistake. We have one of the most fertile periods of innovation that we've seen. When I look at the history of the technology industry, I see periods in which there were low barriers to entry, where there were a lot of small inventors who were not trying to protect their intellectual property rights, and then a period of consolidation in which people start playing by a different set of rules and the level of innovation actually goes down. So, if the purpose of so-called intellectual property rights is to protect innovation, I don't see it working.
So, my basic message to Jeff was actually summed up in a phrase I heard once from Norman Mailer when I was in college. He said, "Being evil is doing something you know is wrong. Being wicked is upping the ante without knowing the consequences." I think that some of the behavior that's going on in our industry right now is wicked. People are changing the rules of an industry that is working really well and I just think it's unwise.
So, that's kind of a different vector than any of this discussion. Once I became involved and started looking into things, I realized that there was a further vector about the Internet culture that was extremely important and that's in the area of prior art. And, some of this is related to what my business is. I've recently come to characterize my business not as being a publisher, but as being somebody who watches what some people refer to as the alpha geeks are doing, and figures out when it's ready to reach the next level of deployment. And then we go talk to those people and get them to write down what they know. Now, the fact is that knowledge is circulating in a fairly large group of people before it's published in a forum that, for example, we might have considered publication even 20 years ago.
At least in the Internet world - I won't speak about other technologies - the people who are using many of the key technologies that are shaping the future, are, in fact, communicating in media that have not historically been considered publication and yet certainly are succeeding in disseminating the ideas very widely. So, for example, you'll find many technologies in which the teaching, if you like, of the art occurred, on a Usenet news group or an IRC channel or some kind of discussion group or through the dissemination of source code from an FTP archive. And, obviously, if thousands of people are learning the art from that method of publication, that constitutes, at least in my mind, a valid set of prior art. And I think that that's part of the reason why my customers get so hot and bothered about software patents.
Some of it is very much misinformed. The media has hyped the issue up. People hear a broad, general description of a patent. As Todd is fond of pointing out, they haven't read the claims, they haven't actually read the patent and yet they react in a knee-jerk way to the broad characterization of it. But, at the same time, there is a real germ of truth to that anger, which I would love to have the people who are involved in the patent industry listen to. And that is, this whole question about whether a patent is obvious, is really more rooted in a change in the way that information is disseminated in the information age. And when I look into how does the patent office search, where do they search, it seems quite clear to me how hard it is nowadays in this industry in transition to find out what it is that people are doing.
Now, let me give a very concrete example. I didn't actually file a patent, but by the standards of patenting today I could say that I "invented" Web advertising. I was the first person ever to do an advertisement on the World Wide Web. And, the fact is I know that some small number of people probably know it. But a few years ago, Wired Magazine claimed that they were the first people to do it. They did it in 1995, I did it in 1993. Now, here's the question that really hit me about this prior art issue: How difficult it would be to find that? Even though I knew it, it took me a few weeks of digging through old boxes of files till I could find a copy of a brochure from GNN, the Global Network Navigator, from 1993 that showed one of those ads. Our Web pages are long gone, maybe you'd find it in a Wall Street Journal article that appeared at the time, but maybe you wouldn't.
The fact is, it's very, very hard in an industry that's moving as fast as we are, where, for example, the Usenet, (which is the source, I believe, of a great deal of prior art in many of the Internet technologies) the archives are no longer even available. And so we have a problem which we as an industry need to come to grips with in order to help the patent office to do a better job.
We're not figuring out how to provide the information that we learn our skills from in a form that is easy for them - for people who want to file patents that are valid - you know what I mean, Jay makes the comment, what good is a patent filing that isn't valid; if you don't have a proper title, you have a problem. And, so one of the things I'd really like to explore are mechanisms for helping patent holders, helping people who are threatened by patents, and the patent office to find prior art in this radically changing, rapidly changing industry that disseminates its information in ways that are not traditional, paper-based methods.
Vint: Let me ask if we could stop there --
O'Reilly: That was my last line.
Vint: All right, and go on now to Larry. Thank you.