While I think it somewhat reckless to assume that someone who calls himself a system designer works a great deal with computers, in my case that's accurate.

So I'm going to tell you why, where I see things going, and something I feel in urgent need of addressing in the meantime.

A computer is a type of tool that will do anything I say as long as I can precisely articulate it.

It accomplishes this by interpreting the words I use into symbols, which eventually become actions, through a process called computation.

The language I use to speak to these machines is known as software.

Understanding this language enables me to create real, novel structures of dazzling complexity, poetic elegance and infinite stamina, with no more labour than it would take to write a story imagining the same thing.

And much like any other good story, my beliefs, style, indulgences and pathologies will make their way into my work one way or another.

So in order to see positive results, I must be able to say what I think, and my thoughts must be clear.

A certain conceptual integrity is essential for any maker of things — but for software, there is no difference between what we say and what happens.

This is a deeper literacy than an understanding of current technology.

I believe it to be an essential component of a future that has already begun, and this is why:


The bulk of our industrial computation is currently performed by digital, electronic computers.

The extent of the real-world effects they can muster are weak electrical pulses.

This turns out to be more than sufficient for the manufacturers of electronic devices.

These manufacturers can, instead of designing a system from scratch, tie a bunch of low-cost components together with a small computer and fill in the blanks with software.

We can safely say this has revolutionized the manufacturing industry.

It is also barely the beginning.

Researchers are hard at work putting computation directly into materials, enabling us to program physical forms with no more effort than it takes to program a computer.

This will enable ordinary people to literally will objects into existence.

This technology is expected to be market-ready in about 20 years.


I may be overreacting but I interpret this as pretty significant.

How important will it become to possess things?

What would it mean to save up for a new home theatre or even a car if you can just download the plans and make one?

Imagine the 3D printers of today, but what comes out of them actually works.

Will we see the industrial design equivalent of bloggers?

Will we see Etsy members competing with Alessi? With Apple?

Will we see open-source kitchen utensils?

Will the manufactured goods industry go the way of today's music industry? Or how about today's newspaper industry?

Will we try to build houses or bridges this way?

When I consider these things, I feel anxious — but probably not the way you might expect.

What causes me anxiety is the way I see us treating a great deal of the business of creative work, including but not limited to software.


The commerce I observe around much (but mind you not all) bespoke creative work goes something like this:

A vendor agrees to deliver to a client an artifact of the client's specification on a predetermined date.

In return, the client will pay the vendor a sum of money, sometimes a portion in advance, and the rest upon accepting the artifact.

Often the scope is ill-defined and the acceptance criteria are flimsy or subjective.

Sometimes the artifact isn't even appropriate for the client; it won't solve their problem.

From this angle, it looks like the focus is on a thing, not what that thing can do.

It seems a lot like committing to a result before solving the underlying problem.

And it also seems like a bet that I would have a slim chance of winning.


In my experience the very act of making this bet seems to work against the ability to win it, as unrealistic as it is in the first place.

Furthermore, it doesn't matter if no one believes it will happen, because the effects appear to be the same.

First, the goal ceases to be solve the client's problem and becomes get off the hook.

And the solution to that problem typically isn't very elegant.

As such, we don't ever discover the value of the solution to the original problem, because nobody actually solves it.

Neither do we establish the value of the time of the people solving it, which often saturates as the deadline approaches.

And this is the situation that troubles me, because I see a lot of work done this way.

And I wonder about the things in our lives that we depend on, and if they have been made this way.

And I wonder about the things of the future.


I see permutations of this condition affecting creative professionals from logo designers to the founders of venture-backed software startups.

The solution I propose is not a new one — it begins with splitting the deal in two.

The first deal is to understand and solve the problem, billed as a function of the time of the people solving it.

In my experience, one person is often enough.

It's important to recognize that although this process isn't capital-intensive, we can't say up front how long it will take.

The second deal is to execute the solution, billed whichever way makes sense to both parties.

Now, I have my doubts about a minimum size for this approach if the goal is actually to solve the problem

Because if the job is truly trivial, it won't take much time — on the meter — to confirm it.


I want to close with one more fundamental technical concept: the algorithm.

An algorithm is an effective method for carrying out the solution of a problem.

An algorithm will always return the same results for a given set of inputs, and do so in a predictable period of time.

A cake recipe is a kind of algorithm. If you use the same ingredients and tools, follow the same steps, and bake for the same time at the same temperature, you will always get the same cake.


In the 1940s, the mathematician George Polya wrote a book called How to Solve It, in which he prescribed the following method:

Understand the problem, devise a plan, execute the plan and examine the result.

I submit that if that works for a formal system such as mathematics, it ought to work for even the most trivial creative venture, because in my opinion, creative work begins with problem-solving.

I believe that if we decouple our creative business ventures into first understanding the problem and defining an algorithm to carry out the solution, we will enjoy greater success all around.

I've set out recently to define a viable business model around this principle, and I've set up a space where those interested can follow along —

With the goal that customers enjoy the full value of their problems actually solved, and creative professionals can prosper now, as well as in the future.

Thank you for your attention.