Robot Marketing or Human Marketing: Spot the Difference
5 Ways to Test Your Robot IQ
You're a savvy marketer. You understand that AI is reshaping the advertising and marketing industries. But can you tell the difference between robot and human ad creative, copywriting and research in your everyday? Let's find out.
Check out the five scenarios below.
But before you do, strengthen your robot IQ. Sign up for The Visionary, GumGum's curated AI news source.
Can a robot decipher a creative brief’s instructions about insights, audience and tone?
At this point, AI is becoming an essential tool for media buying and campaign optimization. It can also take a handful of pre-selected elements and create dynamic ads that can match a site’s audience or aesthetics. But creating persuasive creative? Not so much. Copywriters and creative directors will still be around to come up up with taglines, brand identities, logos and videos for some time.
Today, AI marketing platforms such as Albert pore over mountains of data in real time and optimize campaigns accordingly, making media buys and even creating different ads on the fly.
“You’ll still need to upload images and videos and text that are relevant to your business, your slogans or promotions that you’re trying to run, or maybe seven different versions of the promotion because you want to test them out,” says Tomer Naveh, CTO of Albert. “Albert will take those base materials and mix and match them according to predefined rules so that, of course, they would make sense.”
So what about art and music? AI has the potential to transform those as well. Researchers are using neural networks to create art—by analyzing sets of songs or a trove of paintings, an AI system can build similar sounds or images. A few years ago, robo-art might have looked like connect-the-dots. Today, they’re much more sophisticated. In the future, AI tools might just be your freelance illustrator or soundtrack scorer.
Want to find out more about AI's impact on the creative world? The Visionary, GumGum's hub for what's next in AI and computer vision, is your source.
That recap of last night’s baseball game you just pulled off a newsfeed: Was it written by an experienced reporter or a robot?
AI tools now exist that can automatically write straightforward text incorporating basic facts and phrasing. Already, the AP, The Washington Post and a few others use natural language generation software to produce basic news stories about sports, elections and business.
Robo-articles are effective for the most functional sorts of news and announcements—they’re almost like a first draft of a story. Ultimately, they can free up reporters to work on bigger, more analytical narratives.
For basic business copy, robo-writing is also making headway. A recent study from Gartner estimated that 20 percent of all business writing will be done by machines by 2018. Narrative Science’s Quill, for example, is able to generate functional real estate descriptions, catalog entries and website copy, while Articoolo is aiming to provide software that can write an article based on any request you make of it.
Results, so far, have been mixed. Text may be correct, but stylistically, AI-generated articles often resemble a list of bullet points strung together with filler words.
Then again, couldn’t that simplify something like user manuals?
When you’re sifting through reams of research, are you easily able to come up with the right datapoint to make your case?
AI’s strength is automating repetitive tasks so it can find things that humans might struggle to locate. Sound like your latest research project?
“AI can take on a lot of the research and reporting tasks, finding the right information, data, keywords,” explains Sam Slaughter, VP of content at Contently. “These are things that computers do better than people."
In the legal industry—where locating a past decision or precedent is often essential to a successful case or filing—so-called robot lawyers may become essential algorithm-based assistants. A high-profile startup called Ross Intelligence uses IBM Watson technology to accomplish legal research tasks. It sifts through thousands of legal documents to bolster a firm’s cases, doing tasks that used to be the domain of fresh-out-of-law-school attorneys. It discovers relevant passages and uses machine learning to continually fine-tune its research methodology.
It’s safe to imagine this kind of robo-research making its way into other industries, including marketing. After all, AI tools are currently being used to analyze large troves of social media data to locate insights about products and audiences.
Note: This copy was written and edited by a human.
Your marketing team needs images that align with the company’s new branding program. What do you do?
Your first thought might be to put together a team of interns and start hitting up every stock image site. Or you could go to a service like EyeEm, which uses machine learning to train its algorithms on the aesthetics of an image, finding ones that meet pre-defined guidelines while weeding out those that are off-brand.
That’s what Boston Consulting Group did. EyeEm created an image search filter that would only find images appropriate to its brand. BCG’s 8,000-plus consultants—not people necessarily trained in visual brand guidelines—use it to find collateral for presentations and client meetings.
“Not every consultant, as smart as they are in running businesses and other things, has the right understanding of aesthetics that might be aligned with the brand,” says EyeEm cofounder Lorenz Aschoff.
To build a filter, a client like BCG provides EyeEm with a range of “on-brand” images, which are used to train the network so it can deliver visuals with a similar aesthetic.
Who do you think will train AI systems to do what they do? If you answered “humans,” give yourself a point.
While coding, analysis and other hard skills are going to be necessary in the age of AI, softer skills like empathy and curiosity will be equally important. After all, AI doesn’t have emotions. So having a high EQ (emotional quotient) may be as crucial as having a high RQ.
“The skills needed to succeed in today’s world and the future are curiosity, creativity, taking initiative, multi-disciplinary thinking and empathy,” explained Tiffany Schlain, founder of the Webby Awards, in response to a recent Pew Research Center jobs study. “These skills, interestingly, are the skills specific to human beings that machines and robots cannot do.”
The human touch is essential for AI success. Think of it this way: Nobody likes a tone-deaf chatbot or robo-caller. Maybe that’s why poets and screenwriters are joining tech firms to script the interactions for virtual assistants like Alexa, Siri or Cortana.
According to the Forrester Future of Jobs, 2027 Report, Autodesk hired a fiction writer to create an “authentic, respectful engagement using appropriate voice and style” for its new IBM Watson-powered chatbots. And it isn’t just writing. Forrester predicts that design and performance—whether it’s a jingle or an annual presentation for media buyers—will still be human-created.
Illustrations by Carra Sykes