| Home | About Edge| Features | Edge Editions | Press | Reality Club | Third Culture | Digerati | Edge:Feed | Edge Search |

The Chef: Nathan Myhrvold

Lew Tucker, Steve Lohr, Esther Dyson, and Phil Leggiere on The Chef: Nathan Myhrvold

From: Lew Tucker
Date: 1-23-97

Since Myhrvold is in general forward thinking in his views, it is hard to understand his "something of a curmudgeon" stance on Java. He correctly points out that researchers in computer science periodically proclaim some new programming language as being the one which will free us all from the drudgery of actually programming computers, but by focusing on Java the language, Myhrvold may be missing what has so many people excited and why it is being embraced by the industry.

Java is not just another language, but needs to be thought of as a software platform that directly addresses issues that are important for the Internet: dynamic network loading of functionality, program safety and verification, security management, and most importantly, the ability to run a Java program on practically anyone's operating system. This is the "ubiquity" play that allowed HTML to have such a dramatic impact - transforming the pre-Web Internet into what we have today. Java isn't the first attempt to address these issues - virtual machines and portable byte codes have been done before. But by bringing it all together into a single software platform and getting it adopted by the major players, remarkable things can begin to happen. No guarantees, but each new twist causes us to think about problems slightly differently and that's what can lead to fundamental changes in the way we work and use the systems around us.

As Myhrvold says, "the interesting thing about Java will be the programs people create in Java". So what will these programs do and how will Java effect what gets developed? When I was a research director at Thinking Machines we were faced with the problem of how to use massively parallel machines to solve computational problems. Over time we became pretty good at exposing the inherent parallelism in each problem because we weren't afraid to think of the solution in terms of having available millions of "virtual" processors . This often led to elegant solutions that were efficient even on only moderately parallel or serial machines.

With the Internet I believe we are facing a similar situation in that we need to stop thinking about desktops and servers and to start thinking about writing software for the Net. What then will it mean to "program the Net"? I don't really know myself, but judging from the number of Java developers itching to develop applications and services that are designed to live in the Net and care little about the underlying hardware system, I'm sure it won't be long before we find out.

From: Steve Lohr
Date: 1-23-97

Nathan Myhrvold is a futurist with a nice sense of historical analogy. But I've often wondered if there isn't something basically flawed in the television analogy, whether cable or broadcast, that Myhrvold mentions in the interview, and Ted Leonsis of AOL always invokes.

The big guys in the on-line world, like AOL and MSN, are building large audiences and, once the user is inside the tent, they try to offer something for everyone. It's sort of a mixture of the old world- TV-style huge audiences - and a new one - the special-interest offerings so suited to computer networks. But in the long run, I don't really see the benefit of the so-called aggregation model (the AOL approach). The ease-of-use benefit is a real one, but it should diminish over time, unless the Net is never going to be the big deal as a consumer medium that everyone's betting it will be. The only other reason for aggregation, it seems, is the corporate one (we're big, so a business has to be big to matter to us).

I'm probably wrong, but it seems to me that the content side of the Net may not be a field in which big corporations have a real advantage. Unless, of course, the "pushers" (speaking of the technology) prevail, and the Net becomes television - no analogy required.

From: Esther Dyson
To: Nathan Myhrvold

Date: 1-25-97

Hey Nathan! You missed the point! It was the East-West High-Tech Forum; now it's "the High-Tech Forum in Europe." The point was always to integrate Europe, not to segregate the East Europeans off in the corner. The fact that this year we held it in (the former) Western Europe for the first time was a sign of success -- and a great source of satisfaction for me. Next year : Amsterdam! (We'd love to have you there!)

But yes, I've mis-recognized some patterns, and learned from it. The best way to find edges is to fall over them!


From: Phil Leggiere
Date: 2-3-97

Nathan Myhrvold's clear-headed and long term perspective on the economics of the web is a refreshing antidote to the all or nothing, instant boom or total bust mind set the topic's usually approached with. His description of the Web as a realm of content blindly chasing elusive customers with no evident profit formula ( the build it and they will come fallacy) accurately dramatizes one powerful dynamic.

I think there's been another, subtler, ultimately just as important process at work over the Web's first 18-24 months of pop cultural life, however, that of "customers" ( not the mass market those looking for blockbuster internet broadcasting success expect(ed), but a steadily increasing critical mass of regular Web users)in search of steady, reliable Web sites to integrate into their overall daily media mix. I know from my own experience and that of most veteran Web users I know that once the novelty of random surfing and the cornucopian glut of information wears off the next step is to critically and selectively organize one's limited Web time, choosing the few sites in each area of interest that really "pay off" from a consumer's point of view and reward regular visits.

The shake- out that'll probably prove more important than the mood swings of Wall St. will be the one in which the first wave of two or three dozen serious pioneer sites firmly establish themselves via critical reputation as a grade apart from the scads of ephemeral, coterie and vanity sites. I think this process of establishing widely accepted hierarchies of popularity and proven quality, though far slower than the other curves Myhrvold discusses, is actually moving along more quickly than we may think.

Phil Leggiere

Back to The Chef: Nathan Myhrvold

| Home | About Edge| Features | Edge Editions | Press | Reality Club | Third Culture | Digerati | Edge:Feed | Edge Search |

| Top |