From neural nets to autonomous machine learning algorithms: The many different w
Artificial intelligence is one of the marketing industry’s principal buzzwords right now. Like many en vogue terms, definitions can range all over the map. We asked eight agency executives to give a clear and concise definition of AI.
Source: Ilyse Liffreing
Jason Goldberg, svp of commerce and content, SapientRazorfish
AI are computer systems able to perform tasks that normally require human intelligence. Machine learning is a subset of AI, and deep learning is a subset of machine learning. Essentially in deep learning, instead of writing a set of programmatic rules, a model can be developed that can be trained with a data set that allows the model “to learn.” For example, I can train a model with consumer behaviors to learn what offerings and services a new shopper is most likely to want.
Robbee Minicola, president of Wunderman Seattle and global lead of Wunderman AI Services
AI are the insights derived from autonomous machine learning algorithms. Most people interchange Machine Learning and AI thinking they are the same thing. You can have machine learning without AI, but you cannot have AI without machine learning. Machine Learning is not entirely self-learning; it is very much bootstrapped to human control. AI has a degree of autonomy that allows the machine learning algorithm more freedom as it cascades over the data. This allows for the discovery of unknown unknowns – and that is where the magic happens. So, where the insights are applicable to business decisions, the result is really just machine learning.
Jenna Niven, creative director, RGA
AI is a bunch of algorithms that are leveraging neural nets to process a huge amount of data. They may be simple like creating word clusters or complex like autonomous cars, but all leverage multiple layers of a network to produce probabilities of a certain outcome.
Jason Beckerman, CEO of Unified
I would define AI as the end outcome. AI takes the data processed from machine learning and makes ongoing decisions about what to do to achieve specific goals. Often when marketers refer to AI, they’re actually talking about machine learning, which is a subset of AI. Machine learning is especially top of mind for marketers, because it enables computers to efficiently analyze data without any manual human assistance, and can automatically dig deeper into insights based on what it learns.
Winston Binch, chief digital officer of Deutsch North America
AI is anything that has the ability to fully replace human intelligence. Most of what exists today are really examples of augmented intelligence as opposed to pure AI. Augmented Intelligence is like lane assist. It’s a partial AI experience. Experiences that are not fully automated and still require some level of human interaction and involvement, but self-driving cars would be AI.
Alexander Rea, creative technology officer of DDB New York
AI is a theory often used to refer to a computer that performs a task usually performed by human intelligence. AI has become a clickbait buzzword that drives impressions. It’s easier to say what AI is not, instead of what it is. Because what it is, is something that not quite known yet. So far, we only have very good pattern recognition, learning from sequences, and machine learning. We do not have any system that is truly intelligent. And if you want to talk about AI and autonomous vehicles, that’s not AI either. That’s simply a computer that knows how to do what it’s programmed to do. It’s not going to know that what you really need is some comfort food and want to go through the drive-thru even though it will make you late for work.
Marc Maleh, global director at Havas Cognitive
AI is a series of APIs that agencies and brands can utilize to create more personalized experiences and content to a hyper-targeted set of consumers. But, if you talk to a university data science professor or someone at Amazon or the NSA, they would say we as an industry are not even scratching the surface of what can be done with this technology and because of this I truly believe that everyone is going to have a different version of the definition.
Jason Herndon, director of technology innovation at Rain
AI is anytime we use technology to solve human problems that in the past, only humans could have solved. The litmus test for AI should include, at some point, a measure of the amount of our lives it’s given back to us.