“The Future of Living with AI” — Insights By Paul DeMott, Chief Technology Officer at Helium SEO

Credit: Credit: Paul DeMott, CTO at Helium SEO

Short-term AI is going to cease to be a consciously used tool and rather a layer of invisibility in our daily decisions. I have already found AI suggesting how to prioritize projects, allocate tasks to staff and even predicting revenue changes with amazing accuracy in the workplace. AI is finding its way into the home beyond voice assistants into less noticeable optimization, like controlling how much energy is used across appliances based on grid demand or anticipating what a person needs in the grocery store based on buying habits. These are not 21st century scenarios. They are in deployment today and in five years they will be as popular as email is today.

The big picture is more complicated. Instead of merely automating what already exists in the industries, AI will change the way industries quantify value. In the medical field, AI will not only scan faster, it will also prescribe treatment with odds of success that rely on millions of previous cases. Finance AI will be more than merely forecasting the market, it will be able to place millions of trades in milliseconds and optimize balance sheets with thousands of variables. The implication of this to society is that the efficiency will increase exponentially and so will the requirement of governments. When it is taking a machine only a few seconds to make decisions that used to take committees of experts, we must determine how we will assign accountability.

The challenge that is most ignored is human rather than technical. Individuals create an overestimation of the willingness of organizations to implement AI in a responsible way. Most leaders are quick to implement models without taking into account quality of data, bias, and the cost expenses of sustaining the infrastructure in the long term. I have witnessed companies pay more than $100,000 in cloud computing within a quarter, and later to find the forecasting power of their model had increased by 4 percent of a mere baseline. AI is powerful, and in the absence of discipline can squander resources at a megalithic level.

Among the skills that I stress on in the next generation of executives are two. The former is statistical literacy, not on the level of a PhD but reasonable to demand that one can doubt that a model prediction is sound. The second one is governance, knowing how to be innovative and accountable at the same time. In Helium SEO we train managers to evaluate the assumptions, data provenance and anticipated accuracy of each model prior to its implementation. This practice will make sure AI is incorporated in a responsible manner, instead of being allowed to go uncontrolled.

It is neither an optimistic nor a threatening future with AI. It is a place where ordinary operations are more and more computerized, where industries are being restructured by anticipatory energy, and where the demarcating boundary between human discernment and machine autonomy is growing unclear. The successful companies will be the ones that integrate technical progress and prudent management, and AI will turn into the tool that will complement human decision making but will not overtake it.

Meet Our Contributor — Paul DeMott
Paul DeMott is an Entrepreneur and Chief Technology Officer at Helium SEO. He’s a passionate digital marketer who writes software and AI to develop better tools for marketing online. Learn more by visiting Helium SEO.