When Does A Designer Become A Reverse-Centaur?

22nd April 2024
5 min read

Do we control the tools we use as designers? Or do they control us?


In analytics and data science, a centaur is defined as the combined advantages of a human and an algorithm to improve decision-making.

The notion is that a weak human and weak machine, coupled by a strong process, is far more effective than either a strong human or a strong machine alone.

Where humans are weak; pattern recognition, semantics, complex calculations, machines are strong.

Where machines are weak; intuition, creativity, interpretation, humans are strong.

A human head controlling the machine body.

The centaur model has been proven to be an effective one.

In a simple example; a clinician looking at spots for signs of skin cancer can call on a suitably-trained algorithm to analyse questionable spots against patterns known to be skin cancers.

The algorithm looks at the spots the clinician tells it to.

A reverse-centaur flips the model.

A machine head controlling a human body.

Using the same example; the algorithm performs a scan of a given patient. The results are then crunched and given to the clinician to verify what the algorithm has identified as risks. This will include any hallucinations is has too - freckles, scars, tattoos.

The clinician has to look at the spots the algorithm tells it to.

However as Cory Doctorow pointed out, humans suck at spotting errors that look right. Trying to identify the difference between what looks right and what is right is difficult, and humans eventually get tired of trying.

The reverse-centaur model only works for whoever owns it (ie. the medical centre).

We are of course, about to talk about AI in design (again).

It's Just A Tool.

Every tool we use as designers to some degree influences our design.

Your sketches will be different depending on whether you use paper or Photoshop. Your block models will be different whether you use blue foam or plywood. Your CAD models too will be different whether you use Solidworks or Rhino.

It's not really a debate. It's just a matter of choosing the right tool for the right job, so it can influence your design in the right way.

A centaur.

But when it comes to ideation, the best tool is no tool. Idea-tion. Pure imagination.

The problem with pure imagination though, is that it's not very good at building complex, detailed, holistic constructs of the thing we imagine.

Or making that portable for communication with others.

We need the tools to help us.

Communication Tools

Something that many designers lose sight of is that all of these tools, everything from paper and pencil right up to CAD models and design drawings, are means of communicating an idea. They aren't the idea itself.

Firstly, they are communication tools to ourselves. They allow us to record our imaginings so we can free up that brain space to think up new things. It allows us to layer one idea on top of another.

Secondly, they are methods of documenting our ideas, in various permutations, to communicate specific information to others for a specific purpose. The purpose of a sketch is different to a design drawing, is different to a rendering, but each has their place in the design process and beyond.

There is a natural hierarchy to these communication tools. As concepts evolve and develop, the macro design, easily worked out with sketches and physical mock-ups, solidifies first. As details get added however, it becomes necessary to switch tools better suited for the purpose.

Using the right tool for the job also communicates to the viewer the purpose of the information being presented. Presenting a sketch communicates that the ideas within are open for discussion and iteration. Presenting a design drawing communicates the idea is ready for production.

More than once in my career we've presented concepts as renderings and the client has been upset that project progressed so far without them. It hadn't, but by looking at highly-polished renders, even if it was only smoke and mirrors, they were convinced the design was finished. Doubly bad if the design misses the mark too.

That's why we start at sketches. It communicates idea and the purpose of the communication.

But to create a sketch, first you must imagine it. Or imagine pieces of it. When ideating, the blank spaces remaining are free to be filled with whatever our imagination can dream up. And because a pencil sketch lacks detail, it also empowers the viewer to use their imagination to fill in the blanks.

As a bonus, when the method of communication lacks detail, it opens the door for miscommunication. And if there's one thing humans are great at, it's miscommunication.

History is littered with breakthroughs made from mistakes.

Which Part is the Head?

Even though the AI tools themselves have significantly improved, I've become more of a sceptic of AI design tools over the last 12 months, rather than less.

I'm yet to see them provide genuine value at the value end of design.

If image synthesis is being used for ideation, which it absolutely is, it is also most certainly influencing the ideation process.

How many hallucinations must be sifted through to find the nugget we already knew we were looking for? How much time must be spent getting these tools to communicate design intent, when a simple sketch and some imagination would do?

When the fidelity of these Midjourney (et al.) concepts are so good (like, photorealistic), what room is left for our imagination to fill in the blanks? What opportunity are you giving clients to fill in the blanks? To collaborate?

At what point does the process become a reverse-centaur?

There is of course no right or wrong answer. What's right for one designer will be different for another. And it's not to say that someday they might actually do what I want them to.

The image for this article was of course created in Krea, after all. I finally managed to find a use for it that I'm willing to hand over control to and be judged on.

ⓒ Lincoln Black 2024

Get some free design thinking, fresh every week. Straight to your inbox.

I respect your privacy and will only send you the fresh design thinking you've asked for.


Virtimachi Pty Ltd

ABN 31 651 760 762

PO Box 7156 Kariong

NSW 2250 Australia