Companies should be about experiences: or how we should invert the cloud
The forces of the industry of scale are hard at work at dropping the price of computing lower and lower. New cloud technologies which allow on demand computing are dropping so low, you’re basically coming to the point where you’re paying only for the energy required to power the processing (the AWS Lambda request costs are calculated in nanodollars). What we’re actually paying for is the energy cost of producing value (useful information), using knowledge as a lever.
In a project a friend and I are working on, we have a goal of matching translations of the whole of human literature and providing it for free. The algorithm is still being tweaked, and we’re working on sourcing the texts, but the amazing thing is that it’s actually doable, for negligible cost.
The evolution of web companies
Internet companies that have been created so far are operating under a principle that has been true up to this point — that the cost of computing (including setting up the whole infrastructure) is very high. Until fairly recently, you needed a huge investment to start anything computing related. Servers needed to be bought, ops hired to manage the servers, database admins to manage the data, multiple developers to build the product, making the barrier to entry high.
The basic principle behind it is that managing customers’ data is hard, and as a byproduct of that, the companies own that data and the customer has a very hard time moving that data out of the company (vendor lock-in).
The companies that managed to succeed (while we’re still accepting the fact that we’re paying for the service with our data) now have huge advantages over new companies coming in, simply because a lot of products become more valuable as more data is available to it.
The inflexion point
I feel that we’re reaching a point where the cost of computation is becoming so low, it’s no longer necessary for consumers to pay for the computing power with our own data. Consumers can own their data, pay for the storage and computing cost associated with working with that data and have the companies bid to access it.
I’m fairly certain that for a few dollars per month you could own your email, images, social posts, medical data, banking data and allow companies which provide the most utility from that data access to it (and revoke access and move to a competitor with a mouse click). The entities don’t have to be only people, a band can own their songs, restaurants can own their menus for food delivery services, etc.
Another benefit could be that the companies could charge you directly (or a percentage on top of the computing cost) for providing that service, removing the business model of “just putting ads on it,” and making them extremely focused on providing you the best possible experience with that data.
A different programming model
From the implementation perspective, companies would become their code. I feel some functional programming paradigm would be necessary to achieve this, with code spreading across datacenters, crunching, aggregating and computing data, and most importantly, using that data to provide a valuable experience.
Who could pull this off?
Even though the actual data could be distributed, and people could even self-host their data, developing this whole new paradigm would require a huge amount of work, but the possible rewards could be enormous. A company providing this service would actually have a “computation tax” on the whole world, putting them in the position oil companies had so far in the industrialisation era.
Google, Facebook, and Amazon are the first companies that come to mind because of their scale and resources available, but I would actually like to see Dropbox tackle this problem, as this looks like a natural progression of their model so far.
