Convener: J Chris Anderson http://jchrisa.net
We started by introducing ourselves. We had a lot of people from the tech side (back end analytics, to web developers). We also had a few analysts and business / VC folks. Good crowd.
Then I handed everyone a sheet of yellow paper and and sheet of green paper. The instructions were to write drawbacks or warning signs about the cloud on the yellow paper, and good things, or “the promise of the cloud” / what users want from the cloud, on the green paper.
Here’s the notes everyone wrote:
The main limits of the cloud were:
* Few players owning everything
* User trust of cloud operators
* User lock-in to particular services
* Network unreliability / latency
* Non-portability of applications on a particular platform
* No single identity model
The biggest promise of the cloud was:
* Easy and fun communication
* Standard interface (the web browser)
* Scalable programming models
* Effiecient resource usage (green)
* Users’ data is safe from loss
Then I presented a way of thinking about the cloud that could turn our expectations upside down. Imagine a device that allows you to carry an entire copy of the internet around with you. Then it doesn’t matter if your connectivity is interrupted (you wouldn’t notice). You could also intentionally turn off your connection, so you can focus on what you already know, rather than being constantly interrupted with new information.
The discussion turned to filtering mechanisms for this cloud. Since you can’t actually carry the whole internet in your pocket, how do you know which subset to carry?
There is also a relevant scaling law: the amount of data people are using on a daily basis is growing faster than connection speeds. This means that you’ve got to have some mechanism to work with data in an offline manner, because you can’t always shove it over the wire in an on-demand way.