Humanizing AI for Sentient Foundation

How do you carry a connection to nature and to people, keep the hard analytical edge that's so natural to AI, and still build the feeling of an open-source community, all in one brand?

Sentient. Capable of feeling.

Sentient Foundation is a non-profit developing open and decentralized AGI. Its mission is to keep the future of artificial intelligence transparent, open-source and available to everyone, instead of controlled by a single corporation.

We've been building brands for web3, fintech and crypto for years, and in that time we've run into dozens of projects that, because of their industry, were supposed to look a certain way. Sleek, minimal, a little futuristic. Lately we've been watching the market closely, and people are getting tired of that look. Black background, a technical feeling built from 3D or digital motifs, one bright accent color. You've seen it before, haven't you?

Design has hit a point where people are tired of sprinting into the future. They want things with a little nostalgia in them. The trends that ruled the 2000s are coming back in new clothes: punk grunge, ethereal floral motifs, trinket design, even naive design. All of it is asking for a return to roots. Sentient was the perfect chance to build something genuinely its own, something that steps clearly away from the accepted look of the tech world.

The machine's dream

So how do you smuggle the human into an AI foundation? Let's say it's a matter of perspective.

While I was digging into the subject of the senses, I came across a film about how a newborn sees the world. One paradox fascinated me: in its first days, a newborn has more developed peripheral vision than central vision.

Our brief kept coming back to one phrase, the "machine's dream." The space where memories, emotions and perception begin. Everything started sticking together into one logical whole, and it landed in a single sentence:

We want to show the feeling of a machine dreaming of being born as a human.

Lovely. But let's come back down to earth. We were building a brand that had to be easy to scale and, above all, functional. Using a stack of layered effects in post, we developed the machine's "lens," built on the inversion of how a person perceives the world right after birth. The center is sharp, the periphery stretched and curved, creating a sense of depth in the lens itself.

I designed it so we could apply it to literally any video or still photo in a way that's instantly recognizable and builds a narrative for the brand. The whole thing is finished off with an aperture that imitates the movement of human eyelids, and in the brand it works as a smooth transition between scenes.

Making images and video the easy way

The lens is no good to us, though, if there aren't the right images underneath it. The nostalgia, the human ability to feel, these things are so fleeting they're genuinely hard to picture.

So we analyzed hundreds of photos, worked out a very specific aesthetic, and built a way of making new assets that solved the problem once and for all. With the right AI tools, we could quickly produce new images with full control, in Sentient's specific and, more importantly, consistent style. And that wasn't all, because the same images could be turned into film and placed under the lens we'd built.

At this stage the human aspect became so vivid it was almost overwhelming. Which meant it needed balancing.

The technical side

We'd landed the human aspect, and on its own it still wasn't enough to communicate that this is a foundation building an open-source AI community. We needed to bring in the technical side too: the analytics, the computation, the flexibility, the room for interpretation that AI makes possible.

Every AI is built on data. From data, it connects threads, runs complex calculations, makes an image. I wanted to visualize that process and create the impression of data being assembled into a logical whole, gathering scattered fragments of information, breadcrumbs, into a defined form a person can recognize.

The solution had been in front of us the whole time. We just had to change the thinking we'd brought to it back in proposal two. The "dune" effect from that proposal, the one originally meant to evoke grains of sand, now worked as a filter over the reality a human sees, and became another layer of the AI's perspective. It let us convert the high-resolution images and objects the AI had already generated into an analytical layer. Not only photos, but, even more importantly, film.

Snipped code

It's no secret that the "prompt" is integral to AI. And lucky for us, prompts are strongly associated with monospace typography, partly because a computer reads letters more easily when they're all exactly the same width. We used that relationship too, and brought snipped-code animations into the design of the site.

The color palette

Finally, to communicate Sentient's qualities properly, we based the whole palette on motifs drawn from the natural landscape, pulling from photographs of the sky, meadows, rocks and plants. It gave us a common denominator between the typography, the backgrounds, and the imagery we'd built earlier.

What it was all for

Good branding gives shape to an ideology. Ideologies come in different sizes, some more complicated than others. Lucky for us, this case had to give shape to a genuinely complicated one. An ideology that reframes technology as something that can coexist with us, fully open to us, instead of something we keep chasing until we're doubled over with a stitch in our side.

Something open. To people.

{{case-study}}