A Lifetime of (Machine) Learning

That old adage “never stop learning” is true for people—and computers. Marc Maleh, global director of Havas Cognitive, explains artificial intelligence and machine learning in a way that even his four-year-old can understand. Turns out, the concept may be simpler than we all thought.

Marc MalehTell us about your background.

I started my career as a designer/developer during the dot-com boom (and later bust) after teaching myself how to code in college. I worked for a few startups in New York, building products for clients such as Apple, World Trade Center Redevelopment, and the International Olympic Committee (IOC). Then I enrolled in NYU’s Interactive Telecommunications Program, where I studied computer science and fine art and worked on installations, data-driven art, and robotics projects.

After living and working in Shanghai for a year, I joined R/GA in more of a management role as one of the leads on the Nike account. Later, I spearheaded digital at Wieden+Kennedy and started the data science team, as well as the rapid prototype studio at R/GA. I was proud to join Havas in January.

Tell us about your role at Havas.

My job description has me working across all accounts and offices focusing on machine learning (ML) and artificial intelligence (AI)—and how those technologies can be used more effectively on the creative side of the business. But as we all know, a job description is sometimes just a guide and not written in stone. Since I’ve started at Havas, I have worked on AR/VR projects, as well as Internet of Things (IoT) pitches, and traditional campaigns.

How would you explain cognitive intelligence to an eight-year-old?

This is a very existential question. Having a four-year-old son at home, I have the pleasure of watching his brain grow, learn, solve problems, and reason. All these abilities really define what makes an intelligent solution. This, coupled with the fact that he has no biases right now, would make explaining cognitive intelligence to him fairly easy, I think.

I would have him think about a particular baby toy where he had to put a square peg in a hole, only to realize the hole was round. He made this mistake once or twice, but then realized that it would not work. He then applied that reasoning to other toys. In many ways, you can think of some machine learning systems like that as well.RobotHow do you explain cognitive intelligence to adults?

This is actually more challenging to me since adults are biased due to years of years of hearing and reading about these things, i.e. buzzwords. This is especially true in the advertising industry. A lot of adults would think that AI and ML are replacing them (which in some cases is true) and other adults would think that AI is what you saw in 2001: A Space Odyssey, and that we are no place near a world like that. The reality is that we are someplace in between, and most people use AI systems on a daily basis and don’t even realize or think about it. Those beautifully designed systems that predict traffic patterns and home environment settings, such as Google Maps or Nest Learning Thermostat, function without you even knowing that you are using an AI system.

What’s the difference between cognitive intelligence and emotional intelligence?

I am sure there is a psychologist out there who has a better answer to this question than I do, but emotional intelligence focuses more on a human’s awareness of his or her surroundings based on experiences.

What do marketers need to know about cognitive intelligence?

I would actually take a step back here and think about the modern role of a CMO. A CMO no longer cares about a TV spot, print campaign or banner ad exclusively. Especially when the effectiveness of banners ads is, well, not great.

The CMO’s role is now directly connected to products, and, in many cases, the digital products that a company is responsible for come from the CMO’s desk. Consumers want experiences today; they don’t want to be just messaged to. Gone are the days when a brand could just talk without walking the walk. If a brand is all about energy efficiency, for example, don’t just tell the consumers they should be more energy efficient. Give them the tools they need to be more energy efficient and show them what you, as an organization, are doing to keep that promise. I would argue that most of those types of experiences are driven by data and cognitive solutions. As an agency, we need to be able to explain to clients from all backgrounds (those in marketing and technology) that data and cognitive solutions can and have to be part of their whole ecosystem.

How might cognitive intelligence build ideal relationships between brands and consumers, if there is such a thing?

I think there is such a thing. The relationship needs to be one based on experiences, though. I like to say data in, experience out.

What challenges might there be with cognitive intelligence?

These types of systems rely on data. I think one of the biggest challenges that I see with our clients is sharing data. Not between client and agency but internally, within a client’s business. A lot of times, organizational politics prevent clients from sharing the data and technology resources to be able to go to market with AI-driven experience or campaign.

Another obvious challenge is the ethics around data collection and usage. Consumers are more savvy now than ever and are more protective about where their data goes and what is done with it. As an agency, we need to make sure that we respect this and continue to ensure that all standards and security measures are in place to protect the users. We have a responsibility to ensure that trust between agency, client, and consumer is not broken. As we have seen, when trust around data is broken, brands and agencies suffer immensely.

What does the future of advertising look like?

As an agency, we need to be able to speak to more than just the CMO. The business of running an agency that just relies on just traditional methods of marketing is changing. Consumers are spending less time watching TV in the traditional way, and brands are fighting now more than ever for a consumer’s attention. The future is one where we can create experiences that can drive brand engagement and create platforms for brand messaging. Not just messaging on existing platforms, but thinking about emerging platforms like VR/AR, voice, and chat.

What keeps you up at night?

Other than my kid and puppy?

What advice do you have for those in their 20s and 30s who are at the beginning of their careers?

Two things. One: Don’t be an asshole. Be humble. People have worked their asses off for many years to get to where they are today. Just because you graduated from some fancy design school does not mean you will be a creative director the day after your graduation. Work hard and treat people right.

Two: Be collaborative. Seek out opportunities to work with experienced designers, data scientists, technologists, writers, etc. And request that of your leadership. Agency teams don’t have to work in pairs anymore. They should work in small groups of people with diverse backgrounds and experience. I promise, the work will be better by following that.