Artificial intelligence: A reality check

Artificial intelligence (AI) is the new black, the shiny new object, the answer to every marketer’s prayers and the end to creativity. The recent emergence of AI from arcade halls in academia and back halls in computer science has been spurred on by stories of drones, robots and driverless cars made by tech giants like Amazon. Google and Tesla. But the hype transcends daily reality.

AI has a 50-year history of mathematical and computer science development, experimentation and thought. It’s not an overnight sensation. What makes it exciting is the confluence of big data sets, improved platforms and software, faster and more robust processing capabilities, and a growing cadre of data scientists eager to leverage a wider range of applications. The prosaic daily use of artificial intelligence and machine learning will make a bigger difference in the lives of consumers and the brand than the flashy applications asked in the press.

So consider this AI reality check:

Big Data is messy. We create data and connect large datasets at extraordinary prices, which are multiplied each year. The growth of mobile media, social networks, apps, automated personal assistants, wearables, electronic medical records, self-reporting cars and appliances, and the upcoming Internet of Things (IoT) create enormous opportunities and challenges. In most cases, there is significant and lengthy work on adjusting, normalizing, filling and connecting different data long before an analysis can be started.

It is difficult and intrusive to collect, store, filter and connect these bits and bytes to any given person. To put together a so-called “Golden Record” requires considerable computing power, a robust platform, unclear logic or deep learning to connect different data and appropriate privacy protection. It also requires considerable skill in modeling and a quay with computer scientists who are able to see the forest rather than the trees.

One-on-one is still aspiring. The dream of one-on-one personal communication is on the horizon, but still ambitious. The gateways are the need to develop common protocols for identity dissolution, privacy protection, understanding individual sensitivities and permissions, identifying inflection points and a detailed plot of how individual consumers and segments move through time and space in their journey from need to brand preference.

With the help of AI, we are at an early testing and learning stage led by companies in the financial services, telecommunications and retail sectors.

People Prize Predictive Analytics. Amazon trained us to expect personalized recommendations. We grew up online with the notion, “if you liked this, you’d probably like it.” As a result, we expect favorite brands to know us and to responsibly use the data we share, consciously and unconsciously, to make our lives easier, more convenient and better. For consumers, predictable analytics if the content is personally relevant, useful, and perceived as valuable. All that’s missing is SPAM.

But making realistic, practical data-driven predictions is still more art than science. Humans are beings with some predictable patterns of interest and behavior. But we are not necessarily rational, often inconsistent, quick to change our minds or change our course of action and generally idiosyncratic. AI using deep learning techniques in which the algorithm trains itself can go some of the way to make sense of this data by monitoring actions over time, adjusting behavior with observable benchmarks, and assessing deviations.

Spreading the platform. It seems that every tech company is now in the AI ​​space and is making all kinds of demands. With more than 3,500 Martech deals on top of countless legacy systems installed, it’s no wonder marketers are confused and IT guys are stymied. A recent conduct survey revealed that 38 percent of marketers surveyed used 6-10 Martech solutions and another 20 percent used 10-20 solutions. Bringing together a coherent IT landscape in service to marketing goals, finesse to limiting legacy systems and existing software licenses while processing massive data sets is not for the faint of heart. In some cases, AI has to work around installed technology platforms.

Artificial intelligence is valuable and under development. It’s not a silver bullet. It requires a combination of skilled data scientists and a powerful modern platform guided by a customer-centric perspective and a test-and-learn mentality. Operated in this way, AI will deliver much more value to consumers than drones or robots.