Portal project looks back to the future

In 1998, Joe Firmage took the risky step of disclosing a visionary experience that convinced him of a connection between the world's religions and high-tech advances and visitors from outer space.

The furor surrounding Firmage's revelations coincided with the young multimillionaire entrepreneur's resignation from his position as chief executive of USWeb/CKS. Firmage subsequently became associated with Ann Druyan, a respected science writer and the widow of famed astronomer Carl Sagan, and withdrew from public scrutiny to devise what he and Druyan described as a science-based entertainment portal.

On Tuesday, Firmage returns from the high-tech wilderness to the Siggraph computer graphics gathering in San Diego to announce that his company, ManyOne Networks, has acquired San Francisco-based Web start-up Media Machines, a 3D technology company.

The acquisition of a Web3D start-up might suggest to some that Firmage is looking to the future, or what he calls Web 2.0, by recycling ideas and technologies from the Web's past. And VRML (Virtual Reality Modeling Language) graphics aren't the only suggestion of a back-to-the future strategy: Firmage's plan for a next-generation World Wide Web combines the power of affinity portals, the Mozilla browser, Web directories, and 3D Web graphics--a virtual museum of tried and abandoned Web media concepts. He recently sat down for a CNET News.com roundtable discussion with reporters and editors to talk in more detail about his plans.

Q: First question: Why a portal company?
A: At USWeb I had the opportunity to learn the backside stories of an enormous number of Web projects, including intranets, extranets, and anything else you can imagine. At the peak of the company we were working for half of the Fortune 100. So I saw what was working and what was not. I had a cheat sheet, a list I kept in case we ever had the chance to build a Web 2.0.

In 2000, a year and a half after I left USWeb, I kicked off the predecessor to this project, called One Cosmos. I was inspired by a show I saw when I was 12 years old, "Carl Sagan's Cosmos," a 13-hour series. One of the episodes was called the "Encyclopedia Galactica," in which Sagan said that someday we would have access to all of recorded human knowledge right there on our screens. And I said, "You know what? We could actually do that."

By the end of 2000, nobody was investing in anything to do with the Internet. So I purchased back the company's technology from the initial investors and said I would wait for another day. That day came in October 2001, and ever since then we have been on the trajectory to an imminent launch in next few months.

Tell me about the news you're announcing at Siggraph--the acquisition of Media Machines, along with Tony Parisi, who co-authored the VRML (Virtual Reality Modeling Language) and X3D (Extensible 3D) specs.
One of the most strategic turning points in this prelaunch endeavor was to engineer into ManyOne the proper adoption and integration architecture for 3D media on the Web. And it's more than just integration--we want to adopt the domain and become the Linux of 3D browsers. We want to produce an open-source browser that knocks the socks off anything you've ever seen.

You used the term "Web 2.0," by which you mean?
There were problems with Web 1.0. You need a reusable portal infrastructure based on open-source software. There is too much software that is built by hand which is too expensive and too fragile to maintain. You need to solve the rich media problem without running fiber into every home. The corpses that litter Highway 101 in Silicon Valley are a testimony to the gravity of this problem.

There is so much commercialization of the Web it's beginning to affect the efficacy of the medium.

You need simpler and more powerful navigation of the Internet. You need to find a way to help people get their heads around the Web. You need a medium in which the quality of information can be ascertained. Education will not happen unless there's a way for parents, teachers and students to distinguish garbage from high-quality stuff.

We need a format distinction between commercial and noncommercial portals. There is so much commercialization of the Web it's beginning to affect the efficacy of the medium. Without a noncommercial domain, we risk having commercialism affect the architectural efficacy of navigation. For example, look at search. The more paid search listings predominate, the less the quality of information becomes the ranking paradigm. Whether or not it's good for business, it's a bad thing for the intellectual purity and efficacy of the Web.

And, there's got to be a financial model that lets small-audience, world-class portals--

What do you mean by world class?
I mean AOL or MSN-class stuff. We believe we have developed a portal that would let a 1000-subscriber portal have all the basics you would have in that environment but specifically tuned for that group and financially self-sustained. If that's possible, you could have a financially self-sustaining model, where everyone can basically live in a home environment and not be reduced to the lowest common denominator provided by AOL and MSN.

How are you funding this project?
So far, it's been an $11 million effort, partially from a large number of angel investors. We have chosen wisely not to employ venture capital for this endeavor.

I'm not sure I understand yet exactly what you're trying to do.
What we are trying to build here is the PBS of the Web.

Will it be government-funded?
In our case, ManyOne Networks is in the process of becoming wholly owned by a foundation. We will never go public, and we will never be bought. It will be like a public utility, but not government funded. It will be owned and run by a nonprofit foundation.

How else will it be different from what we see on the Web today?
We've found a way to bring these two paradigms together. Succinctly, we have solidly solved the bandwidth problem; what we've done is invent a new caching system for the Mozilla browser that lets you present Xbox-quality experiences on a dial-up modem.

It will be like a public utility, but not government funded. It will be owned and run by a nonprofit foundation.

A lot of people have asked, "What would a Yahoo with rich media look like?" And we have a radically different answer to that question than anyone has come up with to date. Imagine an online, three-dimensional mirror of reality, of everything we can create digitally on Earth, collaboratively built by the world's nonprofit institutions. Imagine that the Cousteau Foundation could put out the oceans portal, NASA the universe, and so on. What if you had 400,000 of these portals?

Are you using the ODP (Open Directory Project)?
We've grafted the ODP structure onto a new ordering paradigm. This is not an alphabetic taxonomy. This is a natural taxonomy where we use nature, as described by science, to describe what's inside, and what is related to what. The ODP was not designed for that, but it had a lot of articulated structure, particularly in areas of human interest. We've grandfathered in the ODP taxonomies.

So what's the role of 3D in this equation?
The big news is the merger of ManyOne and MediaMachines, and Tony Parisi coming on to lead the 3d team. The future ManyOne browser will include (MediaMachines 3D player) Flux, so anyone who incorporates 3D will be able to create portals that are navigable and cacheable. This is the first viable 3D distribution system because it works on any bandwidth.

Help us envision how 3D will act with this system.
Imagine that six months from now you can click on "Earth," click on "Great Barrier Reef," see a shark there, click on the shark portal, and traverse four different 3D environments, seeing the links overlayed from the brightest minds alive independently stewarding the portals.

And how do you make this financially viable, for these groups to be managing the content?
Our business model is based on a couple of key principles. One is that you cannot reasonably expect people to part with money they're not already parting with. So we have struck agreements to private label the dial-up networks of the major telecom players--Qwest, UUnet, Level3 and StarNet. Here is what we can do with that money. Our members pay between zero and $39 per month. It's zero to download the browser and use it. That also lets you see all the 3D spaces in the "discover" mode only. The educational content for this whole system will be free.

For $21.99 per month, you would get e-mail, instant messenger, dial-up Internet access--everything AOL would be providing you with, and the environment. The cost is approximately $39 for broadband, when it's provided. ManyOne has chosen to distribute this system and subscription offer to our partners so we can private-label the entire environment.

So for instance?
Here's a hypothetical example: Let's say the Humane Society is a partner. Their home page with new rich media wonders gets built, and we further allow them to choose which favorite portals appear by default. Fifty percent of gross margin of $22 or $45 per month will be paid to the organization that brings in the subscriber. The rest comes to ManyOne so we can run the infrastructure--e-mail, IM, etc. In other words, the Humane Society of the U.S. could build their own MSN in an afternoon with no up-front cost and will receive 3 to 4 dollars for every subscriber they bring in. If they could convince their members to switch off AOL and save two animals, then the Humane Society would save a couple hundred thousand.

People pushed these affinity Web sites for a long time, and without a whole lot of success. Why will this be different?
There are, in fact, a number of affinity portals that are extremely successful. But this is less like affinity marketing and more like syndication. Think of ManyOne like NBC with a number of affiliates who take the central stream and add to it. We're in syndication. It's a classic two-tier distribution system, like cars and books and VCRs. By intermediating, we drastically cut the cost of member acquisitions. We can take the cost of marketing to the floor. And that's money we're taking out of Super Bowl advertising and dropping CD-ROMs from the sky and channeling it into the very partners who are not only our distribution partners but also our content partners.

But you still have to market yourself.
That is 100 percent right on the money. The question is making the ManyOne logo be a Good Housekeeping seal of approval. Already, we've been doing partner development, and we have not been turned down once to date. We have got to have the pull-through from users who say, "I love this stuff. It's ad-free and it's supporting an organization that I care about."

The final point is that ManyOne itself is in the process of becoming owned by a foundation. This lets us assure our partner network, who do not want to partner with AOL or Microsoft, that we're in this for the long haul. That we're building something that will never go public, can never be bought and whose board of directors includes some pretty incredible people, including Jane Goodall, Maurice Strong and Paul Hawken.

You've written that your partners will be "socially responsible" organizations. Who's going to count as socially responsible?
It's not a very weighty term in that regard. ManyOne invites its partners to consider endorsing the Earth Charter. But we will partner across the board with commercial and noncommercial organizations, so long as they're not antithetical to our values.

Does that include the NRA (National Rifle Association)? OPEC (Organization of the Petroleum Exporting Countries)?
In general, that will be something for the foundation's board of directors to arbitrate. But the NRA is a classic example of an organization that will have to be approached proactively on our part. They're already in our navigation taxonomy.

There are partnership and stewardship partners. Partners can be anybody, as long as they're not the American Nazi movement. That's why we have a board to arbitrate those extreme boundary cases. These navigation portals--we give by invitation the editorial responsibility to one or more generally nonprofit institutions capable of polishing the lens to look through to that area of expertise. Today you have 100 minimum-wage workers at Yahoo. For the NRA? There's going to be a gun-rights advocacy portal. And a gun-control panel. There will be no thing absent from the ManyOne navigation.

Pornography?
Even pornography. We will have a pornography branch on our taxonomy. You can prune it, if you're an educator or a parent. The taxonomy is not intended to be an exclusionary zone. We only prune it where we broach the level of legality. If you wanted to build an editorial board for the Web, it would have to have thousands of institutions simultaneously doing it. That's what we've created a system to create.

The evidence is in to show that people gravitate toward search rather than directory--and yet you're putting this massive editorial effort into what is essentially a directory with a 3D interface. Second, isn't what you're doing replicating what people are already doing with blogs, that is, having individuals maintaining expert portals of their own?
Not really, because although blogs are doing that, what we're doing is providing an organizational paradigm for it. And the best bloggers would be stewards of portals.

As for your first question, why was there a move toward search in the first place? The number one reason is that the directories suck. They're awful. If you ever try to do anything on Yahoo, you see that it's better than any group out there--but that tells you how bad the environment is.

Let's stretch this out to three years from now. Ten to 300 of the major stewards are out there. Then do a search on ManyOne. The results are not going to be ranked by Google, but by the personal decisions of the best minds in the business. When someone types in "general relativity," there are the links someone should get first. The others, which rank equally high on Google, are garbage because they're pseudoscience.

This is starting to sound like a combination of About.com and Ask Jeeves.
This is Ask Jeeves done right, the stewardship model--it underwhelms it to compare it that way, but done right in terms of letting the brightest people, the most experienced, to build the question-and-answer lenses used by search engines, to prioritize links, to build the contexts, to build these domains.

Being an ISP is a rough business.
It's bloody horrible.

And with broadband it's even worse. There are no margins there.
I do not want to be in the bandwidth-providing business. And we are not. We have on the back end the ability to marry our portal and subscriber business to anybody who wishes to supply the services. We're not interested in buying up modem farms. We'll always be outsourcing that. But we can afford to do this because if you look at the budget, there's a marketing and customer acquisition budget, and we've changed who that goes to.

You refer to your new portal sometimes as the "visual Web." How will the blind and visually impaired navigate it?
One of the things that's nice about our navigation system is that it is positively dripping with semantic meaning. So, for example, if a blind individual were to come up to a future version of this, he could say, 'Computer, tell me what is in the solar system,' and the computer is going to know how to respond to that. Try asking Google that. You could say, 'Take me to Earth and tell me what's in Earth by this semantic mapping of a navigation system.' I think speech-based has potential--we haven't even thought about how far that could go...I think it's very rich in potential in that regard.

You've cast your lot with the Mozilla browser just as AOL is withdrawing its support for its development group. Are you at all apprehensive about the future of that software?
We are going to have a new open-source browser integrating the premier X3D rendering system on the Web. Frankly, it is just fine with me about AOL. The more foolish AOL is in casting off the Mozilla group, fine. We are convinced that an open-source browser is the future of the Web, and market share issues are not telling the long-term story of what's going to happen. I think Mozilla getting out of AOL and into the foundation is a wonderful move.

When should we expect to see your browser and service launch?
We're in the second, hopefully final, beta between 2 and 3 weeks from now, and then we'll launch whenever it's ready. My guess is sometime in September. And then I predict that ManyOne will do for the Internet what the Mac and Windows did for MS-DOS.