David Keshavjee translating die neue Graphik to Japanese with Google’s Translate app.
It all started when Google rolled out a new tool for its Translate app. Instead of typing or speaking to translate, Google made it possible to use the phone’s camera and see live augmented reality (AR) translation superimposed in the image feed. Rather endearingly, the Translate app also attempts to match the original image’s colours, fonts and layout. By interpreting the design of the input, it often creates a bastardized version of the original. “When this tool came out, we found it so great. It tries to adapt to the existing design, which leads to bugs. Sometimes, it changes words; sometimes, the app sees typography where there is none.”
“Text Detection” mode with the white background activated.
The humour of hiding and replacing original designs with their own was obviously not lost on Maximage, who included a “Text Detection” mode that censors anything resembling text with red bars. When the mode is combined with another setting that replaces the original image with a white background, the result is not without recalling some of the visual vocabulary used by Willi Kunz – lines and blocks arranged in staircases – although here these elements have come to life. It was the same Kunz who feared that the facility of design software, which had eliminated the need for traditional technical standards and allowed typography to become more experimental, could potentially lead to the loss of the governing principles of communication behind it.¹ Indeed, the visuals created by Maximage’s app could not care less about fundamental principles of typography. To be fair, however, the app is not so much about communicating as it is about experimenting and creating new tools.
1. Willi Kunz, Typography: Macro And Microaesthetics, 2nd edn, Sulgen: Niggli, 2002, p. 5–6.
In fact, the app was not conceived with a single purpose in mind (such as designing the poster, for instance). In Maximage’s studio, a wall was covered with printouts of both AR outputs and gridded squares to see how the app reacts to them. When I wondered if one of them would be the poster, David explained that “at this stage, we don’t have a poster. We have some ideas, of course, but we are just focusing on developing the tool. We need to discover what is possible. We have been playing with the app and looking at the possibilities to create a corporate identity with it.” And when an application is needed, such as animations for the website or even the poster, “it will happen very quickly, the process will be very intuitive.” Before then, there are still some questions that the designers want to clarify. “AR is not related to print, it belongs to the digital world, to social media. How can we translate it for print? It cannot be a simple printout of the app’s output… We are still defining what it could be.” Ultimately, all the visual material will either be designed with the app or conceived with its use in mind. The identity is thus rooted in the digital tool rather than imagined from its endpoint perspective: the method creates the visual instead of the other way around.
Testing the app in the studio: outputs and potential inputs.
Focusing on the tool rather than its application is what gave the designers room for experimentation. “When we decided to develop the app, we wondered what could be a new and conceptually interesting way of using AR. Normally, its use is pretty limited: it recognises a set image or symbol and triggers an animation. Here, what was interesting for us is that the machine tries, in its own clumsy way, to translate any input.” Any compositional rules come from the image. While the app attempts to find and highlight existing grids, its limitations mean that the output often does not fit within a recognisable grid. “The identity also talks about the limits of a technology that is not slick, but that tries to interpret the world.”
The limits are indeed where the novelty lies. As David pointed out, the generated layouts are not those a designer would normally come up with. In fact, some of the most interesting ones come from the two extremes types of input fed in the app – subjects with highly defined grids or none at all. For instance, aiming the app at the painting of a face sends back a volley of disjointed letters that manically fly around, while testing it on Müller-Brockmann’s three-dimensional grids produces a perfectly formed layout.
There is an appeal to generative design and the idea that an unlimited amount of visuals can be deployed from a set of rules. But if a thousand possible solutions can be created, what is the inherent value of each visual? Is it important to stop, pick a possibility and develop it? “Both are interesting. The process of generative design allows getting out of the tools we already have, such as InDesign. Then, within that, we look for what’s interesting in the new shapes. We do not correct them too much: we take into account the mistakes and select results which are promising.” Furthermore, Maximage are not simply providing visuals to the festival. Instead, they are giving access to the AR app to the general public. The designers are tool makers rather than the providers of visual content, and this feels particularly relevant in the age of Tik Tok videos and viral face filters. With this year’s identity, Maximage remain true to their motto, “emotion and technology”: “what we’re providing is the identity of a graphic design festival, which talks about tools and whose main audience is graphic designers. It is fitting to explore processes and present the idea that tools are great gateways to freedom. If the public gets that, we will have achieved what we wanted.”
