An Opinion on Artificial Intelligence

Data collection is highly valued. There are businesses focused on profiting from them. Data analysts and data scientists are the new careers to aim for. It's not just about big data anymore. It's now about really putting them to work, with the ultimate manifestation building what is known as artificial intelligence.

Artificial intelligence (AI) is here and it's real! Investments for AI development are pouring in to and from everywhere. Progress in the past decade or so in this field has been unprecedented. Still, questions remain. Do we know and understand what we are doing? Are we doing it right? I guess the real question is: are we even ready for it?

A Focus on Data

We understand the need to gather, collect, organize, store and access data. Data can be available anywhere. Anything can be data. Even the tiniest breeze from the wings of a flying bee can be data.

We understand the need to define data with schema, semantics and metadata. Data that defines data is just as important as establishing meaning, purpose and reason to the existence of the data itself.

We understand the need to relate data. At its simplest, organizing and indexing data are already ways to relate data. Data by itself is not good enough, so we relate data to each other and link them to other data. Data relationships can turn diverse collections of data into information.

If these all seem familiar, these are really concepts implemented by databases. Databases put data up front and center. Transactional databases, data warehouses and big data solutions are all in to this, just varying in scale. But it's still all about data. If you think about it, focusing on data is actually the easy part. Data is everywhere.

A Focus on Relationships

Databases mostly use data relationships as constraints. A new breed of databases called graph databases put relationships up as first-class citizens. They are the modern manifestations of graphs in graph theory. Graphs don't focus on data. Graphs focus on relationships. They focus on how data is linked to another -- or on how data is unlinked to another. In addition, the relationships themselves can contain attributes that add more layers of information.

By focusing on relationships, new information can be created out of existing data. The source and form of data are irrelevant. What is important is that the collection of data can be linked or related somehow. Understanding these relationships can provide information and intelligence beyond what data alone can provide.

With graph theory, we can understand the need to gather, collect, organize, store, access, define and link relationships. Unlike data though, relationships are not everywhere. Some relationships naturally exist, yes, but some relationships though need to be created -- or discovered.

Putting It All Together...

IMHO, real AI is possible if we can simulate how our brains use data and relationships together. Consider for example how we decide that we should cancel an event because of another factor or a set of factors. Or how we decide that a series of unfortunate events can lead to something good or bad depending on seemingly unrelated factors/events. The way we think is mostly about how we process data and relationships together.

So here's the hypothesis: "Data is a given. Relations complete intelligence." The relations we put around data makes it useful. How well we relate data so that they can be interpreted correctly creates intelligence. Technologically, we are already masters of data. The question is, are we technologically masters of relationships? At this point, it can be proposed that relations should now be prioritized. If you think about it, relationships collection should now be valued more. Imagine a trend on transactional relationships, relationship warehouses and "big relationships", if that makes sense. Data and relationships together can make AI real. We need to master both. We've gone far ahead with data. It's now time to focus on mastering relationships.