Mechanical Contracting & Plumbing January-December 1909

Our relationship with the internet is a turbulent one. We rely on systems of information to connect us with each other, and to know what we want before we do. This intimacy requires the machine to know us, our habits, interests and friends. But when faced with the inner making of such tools we feel uncomfortable, and even moral repugnance. When companies like Facebook and Google ask for more than we expect we often get upset and uncomfortable, but not enough to change our tech habits.

Like a pressure cooker big tech is efficient in brute force cooking, it is dangerous to look in, and there is an inherit risk in running it. We must not disturb it, and we rarely want to look inside, unless we realize there is a problem. When morally questionable things went into our stew (privacy concerns, experience manipulation, attention economy violations) we are left in a moral dissonance: should we look for our meal elsewhere, or level our moral distaste?

What is doubly interesting is the tyranny of this model. We accept it as a necessary evil, and keep on using it while holding our nose, blocking out everything we know to be true. We often complain about the system in question, within the eco system itself. Posting criticism within the belly of the beast.

On Attention Made Efficient

Attention is not a commodity machines are made to serve, it is a byproduct of a chain of events, starting off in the early days of Silicon Valley – more specially in Xerox Parc when Alan Kay was working on the Dynabook, a hybrid of an iPad and a laptop. It was meant to be user friendly, and interact with the user using graphic user interface (GUI), but there was no software architecture that could support his vision (visual interfaces were a radically novel idea in the 70s).

To build his vision he invited Adele Goldberg and Trygve Reenskaug, and together they architected Model View Controller (MVC). The core brilliance of the model was the idea of linking mental model to computational behavior, mediated by a tool. The data is called a model, the user interacts with view, and the controllers run back and forth – parsing information and propagating changes. MVC is very efficient in delivering tabular data to users, and capturing their interest and preferences.

The trio developed MVC long before the internet was open, and before any social network was ideated, yet their invention was a natural fit for the promise the internet was meant to fulfill.

The different components of this architecture are a perfect fit for attention economy, and platform design. Once setup such platform is self sustaining in doing what it is designed to to, and as these components make our system, they are a perfect fit the online business.

There was no way for design pioneers in Xerox Parc to imagine the future we live in today, nor that this architecture will drive experience commerce.

MVC was the obvious choice for businesses looking to get on the internet in 1993, in particular the remote stationary data (akin to a warehouse), and the proprietary interface (website, similar to a store).

To this day, MVC is de-facto the only server architecture we use.

We should model this behavior because this is where the pressure cooker starts. A system running on MVC (i.e. all software) is like a building. There are data servers in the basement, a security guard asking for a password in the front desk.

An authenticated user walks in, her movement gets tracked and optimized – and later be sold to advertisers. Experience captured and monetized, with excess data stored for future use.

Whenever our user needs something from that system she needs to metaphysically travel to that location (opening a site, or an app constituters travel in the experience economy).

This reinforces a lot of the problems we all know so well. How much data are we leaving behind? What aspects of our experience are mediated and which ones are real? Was my experience designed by me, or have I been nudged?

It also blurs definitions of agency, identity and value extraction. Are the (FANG) companies creating value, or simply distributing value created by others? If I am the one traveling the halls of this building, are the decisions I make in my movement generate income?


What if we could shift this paradigm to correlate to the moment of value exchange? There is a world of opportunity waiting for those able to offer an alternative to an incredibly dissatisfied public.

The surprising hero in our technological conundrum is a plumber.

A plumber comes to your house, fixes your pipes, gets paid and leaves. The pipes might be different on every job - that is ok, and the plumber is able to adapt to each constellation. As opposed to the movements in the building, which are limited and deterministic by design.

The plumber must be paid, the value exchange is implicit and direct – there is no ‘free lunch’, which immediately eliminates issues of privacy (plumber would never take your pipes with him), or the ‘pressure cooker’ opaqueness.

~ Realestate Plumbers
utility fixed adaptive
creativity machine bound human bound
privacy necessary evil non issue
value exchange hidden, non implicit direct and immediate
data ownership ‘building’ / business operator user
biz paradigm efficiency seeking scale effective service

The backdrop to this argument is that efficiency is becoming cheaper, and as a result differentiation is harder to achieve, and articulate.

The experience manipulation model is bound to plato in the long run. The public discussion is becoming more intelligible and actionable – and is such in the face of what is essentially the one trick pony of data driven advertising.

There is an exponentially large market segment waiting for new definitions of digital experiences, scaling laws, and product–experience dichotomies.

The future is thoughtful, slow, deep and effective – the fast, and shallow is already automated, and the market is not having any of it.