By Stan Moote
More often than not, we are struggling so much with day-to-day activities we miss the obvious. Many of us create dashboards to carefully watch certain activities within our businesses. These dashboards are typically updated monthly or quarterly for management review meetings. In some cases, dashboards are close to real time, so certain activities can be watched and scrutinised instantly.
People who need to make decisions love data; for that matter, we all love data to help steer us in the proper direction. The skilled executive knows how to weigh the balance between too much or too little data versus to wait or jump in and finalise a direction. With all this talk about artificial intelligence (AI) and machine learning (ML), does it make this decision versus time process easier or harder?
AI conjures up thoughts of robots taking over the world. However, in reality, it helps us sift through huge data sets to correlate or find a specific trend or usage requirement. As people, we provided the data sources and the output requirements. AI is not new; think back to your school days of curve fitting to find an equation based on specific data points. The difference now is big bata coming into the picture. There is so much data and as mere mortals, even with some computing power, we do not have the resources to sift through the data. This is where cloud comes in. By using hundreds of thousands of nodes around the world, the possible potential of raw data analysis is more than just amazing — it is difficult to comprehend.
Analysts and expert consultants are often called in to provide more inputs into the decision-making process. Will they be replaced with AI? Certainly not. They know how to “tease out” the necessary data for specific requirements. This is exactly where people come in.
We use ML and AI to understand some specific trend, automate a task or find out what we do not know, so we can have new inputs into our day-to-day tasks. Ad-placement engines are the perfect example of high computational requirements based on a series of inputs. Years ago these were pretty simple — ad length, budget, audience and spot availability. AI engines were created to correlate this to place millions of spots every month. Now, with so much personal data being collected and the industry going beyond a one-to-many linear model, ad placement is more complex and it is cloud computational technology that makes this work.
There are many more practical uses for AI and ML for our industry beyond ad placement. Start with some simple tasks you do monthly or quarterly. This takes us back to dashboards: Do you need to review all the dashboards, or can you simply train a machine to know what you are looking for? How do you correlate between the different charts and graphs? Then let the machine do the work and provide you with the results. In many cases, you will be able to automate these outputs to react and change workflows automatically, or initially by flagging you before the algorithm reacts.
Seems cool? It is. Does this put you out of a job? Certainly not — this allows your organisation to focus more on how to adapt to changes, trends and needs rather than focusing on day-to-day activities. The bottom line is, we have been using various forms of AI in our industry for years: ad placement, archive, special effects, automation — even adaptive algorithms in compression engines. Now with the immense amount of computing power available, AI can benefit every phase of the content chain, from production and post production to management, monetisation, distribution and delivery.
You need to take the leap and learn how to capitalise on AI and ML to grow your business.