Art. Lebedev Studio has developed technologies that magically bolster human abilities. Dazzling brand identities, seamless wayfinding systems, unforgettable advertising campaigns and intuitive interfaces are now created with the help of bots.
Design at Art. Lebedev Studio is now done by engineers.
Neural networks offer an unlimited number of fresh perspectives and can create thousands of new meanings. Engineers set the parameters and the networks come up with fascinating (and sometimes entirely unexpected) results.
Art. Lebedev Studio developed an AI designer and for over a year was passing it off as a human. To date, the network has been successfully used in over 20 of the studio’s commercial projects.
Humanity has already wasted enough millions of hours on brandbooks and “how to correctly place the logo on documents” illustrations. Engineers can now build bots that churn out pitch-perfect mock-ups in the blink of an eye.
Design can be instantly tested on any number of formats. Create a new logo, then click a button to instantly see how it would look on a coffee cup, a billboard, a t-shirt or a car, on Twitter or on a bank card.
“Just throw the pictures and logos into the system, it will generate variants of trucks for you and we’ll choose the best one.”
The system can modify mock-ups based on the data it receives. It can choose the best layout, the right number of columns and the perfect size of illustrations and typeface. These technologies have long been used on the web, but are unique in graphic design.
Brandbooks can now do more with bots behind them. They can be interactive, produce mock-ups, calculate costs and order prints.
Machines took on 90% of designers’ work. And that’s a good thing.
An engineer designs the system, it generates 500 variants and then helps choose the best one.
Example: Raiffeisenbank card design.
Our advantage in making wayfinding systems is our graph that connects signs, buildings and objects around them into information networks.
“The taxi desk has finally been moved from the second floor to the first. All airport maps and signs will be updated before lunch. We’ll reprint the signs in the evening and start hanging tomorrow”
Maps and wayfinding signs in airports 🛫, parks 🌳, stadiums 🤾♀️, railway stations 🚃 and the metro 🚇 are typeset and filled with content by machines.
Designers have developed a style for the automated layout system and Art. Lebedev Studio’s bots can create house plates for any address.
Example: Moscow address navigation
The contents of each sign are chosen automatically. If a new object appears in a park, the system will automatically update all relevant signage.
Example: VDNKh wayfinding system
Russia’s largest stadium can accommodate up to eighty thousand people. All of them require maps to find their way around. Our bots can create such maps.
Example: Luzhniki stadium wayfinding
Developing information monitors in Russian, English and Chinese while taking into account different types of screens (departures, arrivals, baggage claim) and long city names would take an entire lifetime were you to do it manually.
Example: Sheremetyevo Airport flight information screens
“The Aleksandrovskaya—Ekaterininskaya hub opens July 1. We’ll typeset all signs at the hub from scratch and retypeset them at all other stations.”
“The high rise shouldn’t even be here, it has lower priority than other items”
“The city started sending information on ground transport routes that operate near park entrances. Can you add a block to the signs so that bots can populate it?”
“The high rise won’t appear on this sign because the sign knows it’s located so close to the building that you can actually see it.”
“According to our metal friend here you won’t need to have perfect vision to see the sign comfortably from this distance.”
Machines resize media, tag content and locate mock-ups by keywords.
Designing for retail presents a unique set of challenges. Huge systems need to be kept in working order and a huge number of operations need to be made.
Here we need to update hundreds of package mock-ups, verify endless lists of ingredients and bar codes, translate texts, warn about possible errors and roll out designs to marketplaces.
A large online store is a essentially huge bot that helps buy and sell millions of products. It requires a few dozen designers or just one high-tech solution that would generate an insane number of banners, pictures and descriptions.
Designers used to create each banner manually. Now an engineer designs a banner-producing machine.
Designed to look beautiful with product names of any length, any units and values.
Improving the appearance of a website with the help of a designer is expensive. It makes much more sense to design, tune and occasionally adjust an automated design system.
It’s a construction kit that helps everyone get their own version of the periodic table with the data they need.
Maintaining huge systems requires efforts of dozens of people. Interfaces are an example of such systems. The good news is that they have incredible capacity for automation and contain many pages, scripts and paths that can be calculated mechanically.
Designers only do what they do best: design.
Design is now done by engineers