There are few industries left where robots do not coexist with humans. We like to believe that we are still somehow always ahead of the robots.
Although they may be smarter and faster, human traits such as creativity and empathy are characteristics that machines have only had in science fiction (for now). But if our mind is a result of the process of learning and experimentation, then-theoretically-it might be possible for machines to learn to develop these traits as well. In the advertising industry, we have been using “robots” to offer a representation of tasks that the human being performs in this sector. So, if artificial intelligence is optimized and takes care of marketing activities, does it mean that we will be replaced by robots?
Artificial intelligence makes life easier for a marketer. The companies offer an instant logo design based on this type of intelligence (LogoJoy), robots that write public relations materials (PingGo) or even those that produce the showcases of an agency (Saatchi & Saatchi). Also a recent Coca-Cola advertisement indicates that they want to use bots to create music for ads, write scripts, post a spot on social networks and buy media, implying that the AI ad revolution seems closer to Reality than ever.
The scale of changes that artificial intelligence has brought to digital advertising may seem to be subtracting human workforce. When robots already outnumber humans, all we have left is our knowledge and experience to make us competitive against AI and robotics. Soon, they will be smart enough to learn, adapt and grow as the human mind. Is it finally time for technology to take over our jobs?
Analysts facing Deep Learning algorithms: recognition of unpredictable sales peaks
The data is everywhere, but they are just one more thing to achieve success if used correctly. The data analysis process can be extremely slow and does not promise any results. The more data we have, the more difficult it is for analysts and marketers to understand, process, and draw conclusions. But today, machines can learn to read data, interpret images and videos, or analyse almost any type of data flow to find patterns. By using these tools, the life of a marketer would become simpler.
If you look at the e-commerce industry, there are clearly recognisable sales peaks that humans can predict. To give an example, we know that Black Friday brings an incredible growth in sales (according to RTB House, the campaigns grow more than 100% over other relevant dates). We also know that Tuesdays and Wednesdays have more conversions than other days of the week (even 40% more than Saturday).
This general knowledge helps us to plan advertising campaigns and set specific parameters when we decide to pay more for an impression or prepare special creatives.
But the real audience does not work under these simplistic constraints all the time. Your buying patterns can be extremely specific and combine multiple criteria. That’s why today’s digital marketing is, above all, the “one segment” and ads impact an individual based on their personal interests and desires. This happens in unpredictable sales spikes that humans can not know, such as buying a loved one’s birthday gift or planning a family event.
Humans are not able to notice changes in the behaviour of online shoppers, but robots are not only aware of these patterns, they can be trained to spot them right away.
In custom retargeting, algorithms based on Deep Learning – a highly innovative branch of AI that mimics the human brain – can recognise sales peaks like humans, but they also notice patterns that are difficult to predict and react quickly to achieve better goals. On the other hand, the machines do not sleep, which allows them to observe the market 24 hours a day, 7 days a week and adjust the activities even at the slightest change.
According to RTB House studies, the strongest algorithms are already capable of responding to millions of requests per second, which also includes the complex process of request analysis and bid estimation. This is obviously much more than any human could ever analyse.
These and many more tasks are executed by computers instead of analysts, this gives marketers the time to innovate and grow their brand, instead of worrying about how to analyse the data and make a decision that will influence millions of Customers at the same time.
Media planners vs algorithms: reacting quickly to user needs
The process of creating media plans has not changed substantially in recent years, but the number of indicators that media planners need to analyse has exploded. Today, 2.5 quintillions of bytes of data are produced every day and, according to IDC, less than 0.5% of those data are collected, analysed and used.
Meanwhile, artificial intelligence technology is revolutionising the planning and purchasing process, doing jobs in digital media, as well as in radio and traditional media. The daily activities that form the backbone of any media agency: reports, audits, etc., can be fully automated to help specialists focus on strategy and creativity.
With a paradigm shift in machine intelligence, we can derive the vision and act quickly by expanding data sets that we collect. In custom retargeting, among other things, decisions about products to be displayed in advertisements are usually made in less than 10 milliseconds, faster than the blink of the human eye is needed.
The incorporation of algorithms of Deep learning has allowed analyzing the individuals like individuals, instead of the ordinary segmentation in groups. It allows marketers to buy media that can appeal to a user’s behaviour proactively, rather than reactively. Nor is the question of where a particular ad is placed anymore, but who the banner is displayed. Machines can follow a target group and tailor ads to a user’s behaviour and preferences in an ultra-precise way that humans can not do.
Creative Directors Vs. Algorithms: Production of video ads
Artificial intelligence is also intensifying to address the world of creativity. Such is the case of the Saatchi & Saatchi film, conceived, edited and directed by machines.
McCann Erickson Japan brought artificial intelligence to the next level in a creative battle against the world’s first creative director named AI-CD ?, against human counterpart, creative director Mitsuru Kuramoto. Both were given the task of creating an advertisement spot that people would judge by voting.
Although the robot was able to provide a creative direction for the ads or extract from a tagging database and television ads from years ago, it seemed that humanity had triumphed, since Kuramoto got 54% of the votes compared to His opponent: 46%. But that number is terribly close to being the same, so it could be assumed with the same ease that one day the figures can be exchanged.
For now, this fact shows that, in the creative aspect, human interaction and empathy may be what prevents us from being indistinguishable (and perhaps machine replaceable).
An apocalyptic or encouraging future?
The apocalypse of artificial intelligence is actually a hot topic of literary and philosophical debate. Computer expert Eliezer Yudkowsky imagined the following in an article in the book Global Risks Catastrophic: “It would be possible physically to build a brain that computed a million times faster than a human brain.” If a human mind were thus accelerated, a year Of thought would be achieved every 31 physical seconds in the outside world, and a millennium would fly in eight and a half hours. ”
To add reality to this future, a new PwC report has concluded that 38% of US jobs Will be replaced by robots and artificial intelligence in early 2030.
Time will tell whether artificial intelligence can learn to be even more creative and effective than human minds and how it will influence workplaces. But for now, it will boost the growth of many new jobs, including some completely new categories. According to the World Economic Forum, 65% of children entering elementary school will end up in jobs that currently do not exist: some jobs will be extinguished, others will be created.
What we already know about the marketing industry is that when algorithms are able to learn from data, it makes it easier for brands to understand customers on a larger, more global scale than in separate local entities. What humans do with this new power remains to be seen.