Philip Rosedale: Linden Lab is not the Ford Motors of the Metaverse...it's the GM

Many people have written about the rather unusual counter-culture feel inside Linden Lab®. Perhaps most famous are the “love machine” (a way for employees to praise one another) and the “Tao of Linden” (“transparency and openness”). I have heard Robin Harper, VP of Marketing and Community, called “the only grownup in the room.” That’s consistent with my one visit to the San Francisco office, in which I walked by the meeting room filled with low bean-bag-like chairs and half a dozen coders playing with a Wii, in order to have a detailed discussion with Robin about inworld regulatory policy.

But there is more to managing the development of such a sweeping product as Second Life® than unusual names and bean-bag chairs. After all, organizational structure ultimately has to come down to the nature of the product the firm is building.

What I didn’t expect to hear was an analogy to General Motors (vs. Ford). And, by the way, not the GM that is struggling so much these days, but the one that was a creative powerhouse in the early years of the auto industry.

PHILIP ROSEDALE: I would say that I’m a very innovative, ‘think different’ kind of person. I think that we’ve certainly put a spotlight on the organizational things that we’ve done differently. From the very beginning I was convinced, and I’m still very convinced, that we are faced with a technology problem that’s really substantial. In other words, on operating system Second Life, there are only a few technology projects in the world that have the kind of modular complexity that we have. So as a CEO, as a board member, as an engineering manager, somebody looking at the problem of how you are going to build something like this, you, I think, are faced with the problem that this is an extremely complicated machine with a lot of parts. And this has been encountered with the auto industry, for example, that has dealt with this question over history. You say then, “Well, what’s the right way to manage it? Do you take a really centralized approach? Do you take a really decentralized approach?”

And, in fact, if you look at the auto industry, in the first part of the 20th Century, you saw Ford taking a centralized approach and General Motors taking a decentralized approach. It turned out that the decentralized approach proved to be more efficient from a competitive perspective, because you got more competitive economies, and you got more rapid evolution happening in a decentralized framework. So I always took a very decentralized approach to how we were going to solve problems and how we were going to design and set direction for our product. But I guess what I’m saying is, I think that that really was not that radical. Maybe I was so proud of it that I made it feel more radical than it really was. Certainly the Tao of Linden and the Love Machine are kind of unusual names.

But if you actually look at what we’re doing internally, you see something that is, I think, evolving as a best practice in a lot of these complex technology areas, which is, you don’t know the future that well. You got a lot of people with a lot of different passions and ideas working on a rich set of problems. So what you want to do is, you want to build systems that reward risktaking—that result in good decentralized leadership where you are amortizing your strategy-setting and guidance in real leadership amongst a larger group of people than you normally might be able to get away with. And I think the systems that we built are consistent with that. But, if you’re sitting inside the walls of Linden Lab, I don’t think that you would really be that surprised by the ways that you saw us doing things. I think you’d probably be pleasantly, as an investor or something, you’d probably be happy to see that we have a relatively high level of agility for the size team that we have, about 300 people now. That is, if we identify an opportunity or a technology change that needs to happen or a problem or something like that, we move on it pretty quickly. And I think that the speed with which we’re able to do that is fundamentally given to us by this decentralized approach.

So far, Philip has stuck pretty close to what economists typically emphasize as the benefits of decentralization: the ability to give people power over the areas where they truly do have the best information, as opposed to letting a higher=up make a decision about something that they might well misunderstand.

But at this point, Philip moves on to another benefit of decentralization: the benefit of having lots of little decisions being made, rather than a few big ones.

PHILIP ROSEDALE: And I think, as we’re growing, we’re certainly building centralized solutions and systems where we need to. So, you know, it’s a balancing act, but I think what we’ve done right is, we fundamentally adhere to the idea that it’s a really big hard problem, and you don’t want to weigh--centralize your decision making or put all your eggs in one basket or put all your bullets behind one idea that might fail. I think, as Mark said this morning, we have a responsibility to the community. You know, we’re kind of the first movers in something that’s really big and will be bigger than us, and so we feel this responsibility. And I think the best, most responsible way to approach this stuff is to be decentralized in a way that we solve more problems.

As I said when I published the first post about this interview, I am not a Linden Lab insider, and I would love to get more of an inside take on this. For that, I turn to Wagner James Au, who has spent a great deal of time inside the Halls of Linden. to give his reaction. Au's responses to Rosedale will appear on his blog at New World Notes.

Here is a sample of what to expect from Mr. Au:

There's no doubt that Philip's vision of a decentralized company was
a valuable thing, especially in Linden's early years. At the same
time, I wonder if he's aware of the many downsides that involved, too.
In my observation, much of Second Life's limitations are directly
attributable to a lack of a central and unifying vision. It's why you
see tremendous innovations and features added to the software, while
other important areas have lain dormant, even while it hurt the
world's long term growth. Exhibit A: the confusing user interface
which hasn't been significantly changed in five years, even though
it's the chief culprit to SL's poor retention rate. Why? Because
until very recently, most of Linden's developers weren't interested
enough to work on it."

And Roland Legrand of MixedRealities.com has provided his predictably cogent view on Rosedale's vision.

Return to Rosedale Interview article and list of bloggers.