Do designers dream of electric sheep?

aka, The value of human designers in the era of AI (and no, it’s not about empathy)

Cristina Colosi
4 min readJan 17, 2021

It was 1969 and Philip Dick was writing the novel that have inspired Blade Runner. Speculating on the distinctive aspects between humans and replicants (basically, Artificial Intelligence) he asked:

“Do androids dream of electric sheep?”

Then came 2017, and with no explicit intention to answer, Milton Glaser designed a poster for the School of Visual Arts in New York accompanied by the copy:

“To dream is human; the most, perhaps, important aspect of humankind is the ability to dream.”

P : <Do androids dream of electric sheep?>,  M : <To dream is human>>
Philip & Milton reflecting on the same concept from different points of view.

As AI already beats human world chess champions, gets its art sold at Christie’s, and interferes with democratic elections all over the world — soon to become our colleague at work — it’s time for new answers to the questions: What distinguishes human intelligence from AI? And which roles may the two share in designing our future?


This article is not about why AI won’t steal your job as a designer [for that, go here, or here]. The increased interaction with digital brains, more than an employment crisis, may ignite an identity one. But it’s in time of crisis that we discover our most inner strengths, and what is human is to dream.

Abductive reasoning isn’t (anymore) the answer.

When a computer (Deep Blue) defeated the human world champion at chess it was 1997 and, at first, it was annoying. But after a while, we accepted that computers can compute faster than us. In other words, they’re better at deductive reasoning — the game of applying known rules to an unknown situation to deduce information and decisions.

We may as well have heard that the auction house Christie’s sold an AI-generated painting for almost half a million dollars in 2018. As a natural denial in response to the loss, many humans argued that it wasn’t a pure creative act. Yet, by comparing uncountable times its own attempts at painting to existent human art, a machine extrapolated the rules that govern our appreciation of artifacts — rules that we may not even be aware of. With machine learning, AI conquered as well the space of inductive reasoning, the game of inferring the rules of an unknown situation through experience.

Some of us may guess already where we’re heading: abductive reasoning — the art of making an educated guess, inferring previous knowledge to craft a novel model of interpretation for the new, unknown scenario. For years it has been pointed out as the unreplicable bulwark of human intelligence — but, hey — AI is going to college now, and it’s learning new stuff. All it needed was an implementation of symbolic reasoning through which it could not only explore the space of solutions for a given problem but also reframe the problem itself, by generalizing concepts acquired in other circumstances.
To put it in context to designers, AI is knocking on the first-diamond’s door.

AI knocking on the first-diamond door.
AI taking over the double diamond. Coffe and panic are served.

I can hear some of us saying: “Waaaaait…this is bull poop. The first diamond is all about empathy — no way AI is going to be better than us at that”.
Indeed, Philip Dick himself crafted an empathy-test to unmask replicants among humans. But, you know, to sail the sea of the problem space you don’t need to be empathetic; it’s not necessary that you feel what your user is feeling; what you need is to understand it. Actually, feeling it is like a shortcut (some) humans can use to understand! In this context, empathy may be as well simulated.

Check_Mate .

Let’s forget about chess. We’re playing Risk, and AI has the mission card to conquer the two continents of the double diamond.
I think we may have just lost the game.

The residual agency of human intelligence.

Swipe up your tears and keep reading!

There is no need to feel defeated. AI, at its core, is simply a replication of human intelligence, as this is, literally, how we build it. The algorithms of AI are the translation of the sequence of mental steps that we assume a human intelligence performs. But you know what? An AI doesn’t have our distractions. It doesn’t need to sleep, eat, or mate — it doesn’t even die! It has much less worries than us, and it can invest all its computational power in one task. This comparison is unfair.

And that’s the point, precisely.

An AI doesn't need to sleep, eat, or mate: that is to say, it doesn’t have a wish, or a desire. An AI designing a perfectly complying system wouldn’t (hopefully?) ask itself “is the world I’m building the one I dream of?”

Basically, to be irreplaceable, all humans need to do is dreaming.

So please, humans…dream. Claim your agency and design the world that you dream of.

☞ If you liked this article, why not clapping?
﹆ If you have a comment, why not leaving it?



Cristina Colosi

~3 minutes speculations that bridge Science, Design and Magic