Data Analytics Are the Path to AI

By John Papaevgeniou, Founder and CEO of Relational

Most businesses realize that data analytics can increase sales and boost productivity, but there are also benefits for AI. Analytics require automated access to current data sources and the ability to put information to use quickly. That is exactly what we need for the successful deployment of artificial intelligence. By investing in data analytics, firms will have proven systems ready for future advances in AI.

The Reality of Data Analytics

Unlike much of artificial intelligence, data analytics produce immediate improvements in corporate performance. A McKinsey study showed that payments processors using data analytics were able to reduce operating costs by 15% to 20%. They achieved these cost savings while increasing customer satisfaction by 5% to 10%. McKinsey also found that analytics can reduce fraud losses by at least 3%. These numbers are not as fantastic as some projections we see for AI, but businesses cannot afford to ignore them. Data analytics give today's managers a way to improve real-world outcomes.

The Promise of Artificial Intelligence

The enormous potential of AI led to a fantastic expansion of investment over the last several years. McKinsey estimated the global potential value of AI in banking alone at 200 to 300 billion US dollars per year. According to Deutsche Bank Research, venture capital investment in AI start-ups soared from less than 2 to 24 billion US dollars between 2013 and 2018. AI investment in finance started with hedge funds, high-frequency trading, and other speculative fields. In recent years, AI has moved into banking, insurance, and more conservative parts of the financial industry. However, the change in setting has not changed the speculative mentality. There might be too much focus on trying to find the next revolutionary breakthrough.

Incremental Progress

Human intelligence developed through evolution rather than revolution, and we may need the same approach for artificial intelligence. Much of the actual painstaking work of preparing for AI requires investing in building a corporate data processing infrastructure. We often overlook all the empirical observations and data gathering that go into significant discoveries. Those who make the breakthroughs know the truth. For example, John von Neumann emphasized that Isaac Newton's theories would not have been possible without the detailed observations of Tycho Brahe. Similarly, we cannot expect AI to function effectively without access to well-organized, clean, and current data.

The Data on Data Infrastructure

TDWI surveyed organizations that are actively using technologies like machine learning, as well as those only investigating them. 52% of institutions in the active group cited noisy or dirty data as a significant problem. Actually working with data systems also leads to different priorities. 32% of the active group believed that measuring outcomes should be a top priority, while only 21% of the investigating group agreed. As Bill Gates observed, "I have been struck again and again by how important measurement is to improving the human condition." Developing data analytics gives organizations the ability to make the crucial measurements that they need for AI.

Building the Future

There are many small steps behind every leap forward in technology. Investing in data analytics not only brings proven benefits; it also creates the data infrastructure that we need for AI. AI will not be built in a day, but we can take the first step to building it today with data analytics.

Comments are closed.