The automation of creativity: scary but inevitable
The Drum’s latest documentary, produced in association with Teads, ‘The Automation of Creativity’ explores the role of artificial intelligence (AI) in advertising.
Automation will claim 50% of all jobs in the next 30 years, according to Rice University professor Moshe Vardi, but creativity is impossible to automate, right? Adland will surely escape this robot advance?
Such a binary argument fails to take into account the huge leaps artificial intelligence (AI) and other such technologies are making. Why, when it is being used in film-making, music and even journalism, should advertising avoid the onslaught?
Vardi, who specialises in computational engineering, told the American Association for the Advancement of Science: “We are approaching a time when machines will be able to outperform humans at almost any task.”
Automation is set to transform this industry, and arguably for the better. Already we see in the media space the rise of programmatic trading, with many believing it will become the de facto model in coming years. But is tech also the future of creativity? As datasets get ever bigger and algorithms more intelligent, is it not naive to believe creative direction in advertising will remain a human-only pursuit?
In a quest to understand the role of AI in advertising, The Drum, in partnership with video advertising marketplace Teads, has launched a documentary, the Automation of Creativity, which was shot in Tokyo, London and Amsterdam.
The film explores how AI is beginning to impact the role of human creatives.
Martin Talks, founder of consultancy Matomico and former FCB digital chief, says: “There is a certain feeling in the creative industries that it won’t happen to us. That is not the case. AI is already having an impact on the advertising and creative industries – the robots are most definitely coming.
“The defence to that is really pushing ourselves more on creative, as brands and agencies. Clients need to be bolder and stop asking for big ideas while only funding small ones.”
Take the Cannes Lions festival. This year, the feeling was less that adtech had dominated the conversation as it had the previous year, and more that data-driven creativity, big data and technology are starting to shape conversations.
Among the big winners was J Walter Thompson’s ‘The Next Rembrandt’ – a 3D printed painting consisting of 148m pixels distilled from all 346 of Rembrandt’s paintings. The project saw a group of art historians, material researchers, data scientists and engineers take on what many considered a controversial challenge – teaching a machine to think, act and paint like Rembrandt.
Created for banking client ING with Microsoft, it won 16 Lions including Grands Prix in Cyber and Creative Data, and a prized Innovation Lion.
2016 also saw the arrival of the world’s first artificial creative director, AI-CD ß. Created by McCann Erickson Japan under its Creative Genome Project, it might have been a headline grabber but has already been pressed into commercial action for Mondelēz Japan with a campaign for Clorets Mint Tab.
Shun Matsuzaka, the creative planner behind the initiative, says he was inspired by the thought that the industry – for all its talk of data-driven creativity – felt immune from the rise of AI when other creative industries such as film, music and publishing were actively exploring its potential. “Why are we not doing it in the ad industry?” he asks.
A decade’s worth of award-winning work has been inputted into its database and tagged by hand. The team then created algorithms to set the creative direction, with AI-CD ß tasked to crunch its dataset from client briefs.
A third campaign that got the ad world thinking was a digital poster by M&C Saatchi, Clear Channel and Posterscope for fictional coffee brand Bahio.
Billed the world’s first AI advert, the premise was that the poster could ‘read’ the reactions of its audience and adapt accordingly. Fed an initial ‘gene pool’ of images and copy, with 22 ads created in each generation, at each stage the poster assessed the success, or otherwise, of the ad, identifying winners to iterate in its next generation.
M&C Saatchi’s business director for innovation, Samuel Ellis, explains: “We wanted it to come up with new things that a creative would never write and no client would sign off on.” A computer that comes up with something “silly” that no sentient being would consider could well spark the beginnings of a special idea, he adds.
Yet, in each example, AI has been in the execution of the campaign, part of the ‘big idea’ rather than creating the big idea itself. This is something that most agree is not yet possible, though its ability to streamline and broaden the inspiration process is a valuable boon.
Stephane Xiberras, president and chief creative officer at BETC Paris, created CAI (Creative Artificial Intelligence) in 2009. CAI can be programmed to produce ads by selecting product category, objectives and product benefits, and can randomly generate some 200,000 of them. To overcome the problem of programming versus intuition, the agency designed a “huge dose” of random into the machine to give the impression of creativity.
“When I made CAI, I wanted to show in an absurd way that robots make mediocre ads,” he says.
For him, algorithms and other robots are powerful tools to streamline purchases and track targets, but it’s like “the creativity of a hunter,” he says, “with the hunter being the algorithm and the animal he’s chasing being the consumer”.
Xiberras’ opinion is echoed by Yasuharu Sasaki, executive creative director of Dentsu, Tokyo. “I don’t think creativity could be automated. Automation may be able to make created ideas more attractive, but I do not think that creativity that creates the ideas would be automated – at least not for a while.”
Instead, tech itself is a tool to increase opportunities to exert creativity, he says. “I believe it will be used to enhance the expression of those who have creativity to the limit.”
Dentsu is investing big in new tech and ambitious startup companies that will spark innovation across its group. Dentsu Ventures, formed last year, aims to support businesses and entrepreneurs by providing problem-solving solutions and resources unique to the group. Its first project, Jibo, is a social robot that incorporates voice and emotion recognition, natural language processing, machine learning and expressive display and movement. It will, according to Dentsu, continue to develop into a new type of medium that will change the way information is assimilated, with the potential to function as a communication hub in a smart home.
It is in this wider definition of creativity that IBM’s UK and Ireland head of retail Danny Bagge believes tech – whether AI, the cloud, mobile or big data – will help. “It is actually very liberating,” he says. “Every industry, every process, is getting touched, so why should creativity not ultimately evolve?”
IBM is the creator of Watson, a tech platform that uses natural language processing and machine learning to reveal insights from large amounts of unstructured data. It recently helped edit an issue of The Drum and even predicted the probable winners at the Cannes Lions.
Justin Taylor, the managing director of Teads UK, believes it is an exciting time for the industry and those wanting to embrace the tools of the future.
“We are entering the true age of digital now. 20 years ago we were just getting started, and in another 20 years we will get digital right. Tech such as machine learning is going to create profound changes today, in 12 months from now, in five years’ time... It is scary but it is inevitable.”
This article was first published in the 26 October issue of The Drum.