« Supernova - Google | Main | ReThinking Telecom Panel »

David Isenberg

Nice to reconnect with David here. He's just returned from New Zealand where he was working with Citylink in Wellington a small broadband wireless provider. Yesterday I was asking him about bundling telephone, cable, and broadband access from a consumer perspective. We may have to wait for that answer. So I’m looking forward to his update " Why Stupid is Still Smart". Many moons ago we had a great conversation around his paper "Stupid Networks". My argument then as it would be now.... Can’t we apply this same logic to companies? "Stupid Companies are really Smart".

This turned out to be one of the best talks of the day. Of course I was tuned in before. David had his screensaver playing pictures of NZ as we walked in after lunch. Yet finally we were looking at something tangible. David showed us wire that goes in the ground. It has last mile potential. So small and powerful I can have the equivalent of a city exchange in my house in a few years and they don’t even need to dig up the streets.

But take a look at the VOIP example POCKET PRESSENCE and GLOBAL IP SOUND. I you are like me and already use IM to talk direct internationally, then say goodbye to telecom profits with this one!

DI "Most of the important applications haven't been discovered yet. We are really in the green screen era." Now that's a radical thought about POTS!

KW Stupid Networks remain current. Years after it was first written. David will frame from a network perspective.

Detailed notes -- paraphasing David's comments:

DI. Conduits in the ground are infrastructure. Nice modules etc yet it is not infrastructure. It's POTS is still really important and the future for it really uncertain. Headline TELSTRA in slump as PM stalls sell-off. Telco's all over the world are struggling find themselves in hurt city. Verizon, SBC and Bell South, and Qwest...? Qwest is not an exception. It is the first sign of serious trouble. Customer Calling card from Australia. $10. Called Japan put on hold maybe have a few minutes left. Talked for 45-50 minutes. Still had $7.79 cents left. DISTANCE IS DEAD FOLKS. Except where is being held back by regulation or government intervention. 2.9 cents Aust. to call the US. In Japan the news was standing in subway giving away 12mb adsl modems Yahoo BB has gained one million customers from April 2002 to Sept 2002. First Million. now 1.3m In Japan now there are 5m adsl subscribers. In Japan there are 120k fiber to the home vs 50k in the US. Lots of growth! In Japan this growth is on a steep slope. Cost of customer acquisition makes me cringe. Was an old business model just like with with quest. was good for us folks. The fact regardless of the business model they are snapping them up and using them for $15 per month!

The future has arrived it is not just evenly distributed yet! Small co called Global IP Sound. In Pocket Presence there’s Microsoft pocket PC operating system Vanilla 802.11b all in an internet shrink wrapped telephone application and it sounds better than PSTN. Better than toll quality speech. NO TELCO in this model! When the standard vanilla Internet become more useful than the Telco the cash cow goes away. ? Who then runs the network?

What is SIP? What http did for documents SIP is going to do for phone systems! Picture of the "the intelligent network” the best network today is the stupid network that supplies simply connections but no services, Instead service are created by smart network-enabled products. It becomes the connectivity piece for your application designed for any network application.

New cable fiber options coming for the home. Will be easy to light it at a gigabit! In two or 3 years each of us may be able to have the bandwidth of the company... Benefits of Moore’s law are being held back by the telephone companies. Depreciation for Telco’s are holding us back. 5 years depreciation. Assume the bean counters assuming it will be replaced in 5 years. From bicycle to rocket in a qualitative change. Fourth trajectory required. It doesn't scale to anyone’s law. It becomes a new Engineering effort. This means we have to ride the things that do scale and keep the future network simple and stupid.

The major action is the end-to-end principle. David Read, et all conceived of it earlier. If you can do something at the ends don't do it in the middle. That preserves your options. Internetworking shifts control from the network owner to the end users. The Internet says take the features and put them at the edges. Internet's job says ignore network specific differences. (e.g. call waiting - cool features don't confer advantage to the owner of a network.

Telephony intelligent way wires network - service - vertically integrated. In the end-to-end way it is just a shrink wrapped application in an 802.11 connections. There are now a whole lot of new connections that came along with the network. email ecommerce, audio instant messaging , blogging etc. Most of the important applications haven't been discovered yet. We are really in the green screen era.

Old version new business model. They charged for voice applications. In the Smart Network model the potential for a vibrant marketplace exists around applications, data, voice, and ideal anything. The network layer Internet Protocol is a commons. The big question is what the business operating model in the physical layer designed for anything digital. What do you do with who runs this?? Who is best to run this?

Not the telephone co. Difficult to make the transition from the vertical model to the horizontal model. Sailing to Steam ships. Even though much was shared. The method of propulsion killed all the sailing ship lines. This is a bigger transition.

Will it be the cable co? Maybe in this country. Not elsewhere. They have better options. They unfortunately own the current information paradigm. Not sure them it’s them,

Municipalities? Own fiber netwroks 125 cos.
New kind os cos Stokad??? 3 years to put fiber in Stockholm.

There are politics in the decisions. Remember Goliath Lost.

Ross Mayfield. Trading bandwidth as a monopoly. Rights of use agreements. The basic thing is SDNA highest relative to any industry. It is not prepared to operate leanly , as a utility. Will require brand new co's co's coming in. By offering different services. There is a model there. DI only one.

MR Need for identity - framework for decentralized environment. How can we not have servers with directories of information to propagate the system? DI Server is really at the edge of the network. Don’t need to own the wires to own a server. Just an application running at the edge of the network.

Thought provoking and a useful update; the telecoms are in real trouble.

Recent Comments

My Furl


Creative Commons License
This weblog is licensed under a Creative Commons License.
Powered by
Movable Type 3.32