While technology can vary to a wide range of application, from 3D printing to Genome-Editing technologies, the general public will be directly affected in the short term by Data collection and exploration, Artificial Intelligence (AI), and Automation.
- by Jean Saïfi -Source: Annahar
BEIRUT: Technology has been around since the very first existence of human beings. Inventions made us more efficient but each new development came at a cost, from the agricultural revolution to the industrial revolution, where millions of workers had to adapt or get left behind. With the current and projected technologies, this process continues today at a much faster pace.
While technology can vary to a wide range of application, from 3D printing to Genome-Editing technologies, the general public will be directly affected in the short term by Data collection and exploration, Artificial Intelligence (AI), and Automation.
True is the fact that data collection on its own is not the main game changer, but do not underestimate its effect when coupled with AI where complex analytical and predictive tasks are tackled much faster than humans could ever hope to achieve.
Two types of data mainly exist, structured and unstructured data. Structured data is made of clearly defined types that are easily searchable, like phone numbers, postal codes, city, age, gender, and any other defined info automatically generated by a machine or simply inputted by a human. Unstructured data is simply and by definition, anything else! In other terms, unstructured data can be found in text files, emails, tweets, chat conversations, digital photos, videos, audio, speech, etc.
While 90 percent of all the world digital data was created in the last 3 years, we estimate that around 95 percent of the data in the world is unstructured. The yearly evolution of unstructured data is estimated at 65 percent while the yearly growth of structured data is only 23 percent. No doubt by now that the real treasure is hidden in the unstructured data and the best way to unearth it is by using Artificial Intelligence.
Numbers aside, let us tackle a real-life case. You are in Bali and you suddenly decide to take an early AM hike to the local volcano summit to watch the sunrise from atop the mountain. Your friend manages to snap your perfect picture posing alongside the beautiful landscape, nature, and clouds. Excited about that, you decide to share your picture via Whatsapp and also post it on social media. Besides all the positive comments and likes you are getting, a machine learning algorithm (field of Artificial Intelligence) is already running and processing your digital image, applying forms of detection principles and other statistical techniques to both identify and self-learn the different objects in your shared picture. A few seconds later, the AI has already detected your sunglasses shape and model, your hiking shoes brand and their green color, your interest of carrying a bag as well as other information related to what you are wearing. Surprised no? Well, that’s not all. The AI machine will also see the clouds, the trees, and the little-unseen monkeys in the background. So what does all this mean?! It simply implies that a simple picture in a remote area is now converted into a potential business lead, by converting the unstructured data in the picture into structured data, highlighting your interest in sunglasses, green color, shoes and clothes style and brands, travel and love of nature.
After a long trip, you finally land in your hometown. Your first reflex is obviously to turn on your mobile phone (the plane is still in taxi mode) and assure your loved ones that you landed safely. You are now ready to leave the airport but face an endless traffic jam to get home. That’s great news, especially for the online retailers who were getting ready to get in touch with you when the moment is right. The AI machine can estimate that you are now back in your hometown and combines this information with the already collected data related to your demographics (age, gender, income, marital status, etc.) and behavior (historical data and past purchases, payment methods, use of social media, etc.). This will allow it to predict the next best action for getting in touch with you over the following few days. The goal of the next best action can vary from creating awareness about a product, generating interest in the product, calculating the probability that you will ultimately purchase the product, and finally increasing loyalty with the brand and related online sites for future potential sales.
Still carrying the memories of your last trip home from the airport after a long working day, while sitting comfortably in your home couch, you decide to complete the online purchase especially that a competitive price is offered. This also allowed you to avoid another trip to the mall or the high shopping street in the heavy PM traffic amplified by the start of the winter season. You’ve just added another behavioral event to your data and the cycle is only starting.
A few months later, you suddenly realize that most of your purchases are now completed online, from clothing, shoes, electronics, cosmetics, body care, bags and accessories to food and drinks without the hassle of going to the specialty shops and spending an enormous time in traffic and in stores. You are not the only one!
In the Lebanese market, we are currently witnessing some shy attempts from already established companies to create and develop an e-commerce or simply to have an online presence (when you have physical shops and an online store). The only problem is that local online stores are not just competing with each other but also with the big online giants like Amazon and AliExpress. The effect of globalization, supply chains and advancement in communication makes it impossible for local online businesses to match the selling prices and offers listed on the big retailers’ sites, while the consumer today easily has the option to compare prices and select the online store of his choice.
Shopping malls and street shops in Lebanon will gradually start to feel this effect, this in addition to already high rental rates (especially for prime locations) as well as other overhead expenses.
We’ve observed in recent years a linear and steady growth in global online commerce. The percentage of online transactions vs total sales varies widely from a country to another but it is definitely on the rise. The effect on local stores could be massive and the market will need to adapt before it’s too late. Will local shops negotiate lower rents? Will marketing events and activities help increase foot traffic in the long term? Are big shopping malls more resilient than local shops to the online shopping effect? Will Lebanon’s current deteriorated infrastructure contributes positively or negatively in the growth of online shopping on the national level?
The application of Artificial Intelligence and data mining extends to multiple domains. To name a few, hospital and doctors can use AI for a second opinion and disease diagnosis. In finance, an AI can analyze in real time the news related to the different stocks in a portfolio to get the overall sentiment of the market. In customer service, ChatBots can simulate a conversation with human users and execute transactions, understanding speech and typed text while detecting different emotions from the voice tones or inputted text of the human user and adapting to the situation. The effect on the workforce is still uncertain but the question is: will we adapt or will we get left behind?
Jean Saïfi completed an engineering degree at ESIB, USJ and a masters in management at SKEMA business school. He is an expert in innovation and strategy as well as a lecturer at the faculty of engineering, USJ.